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 |
|---|---|---|---|---|---|
China used to be among the countries with a high prevalence of trachoma . At the launch of The Global Elimination of Trachoma ( GET ) 2020 campaign by the World Health Organization ( WHO ) in 1996 , China was placed on the list of countries endemic for trachoma based on historical data . However , empirical observation and routinely collected eye care data were suggesting that trachoma was no longer a public health problem . To determine whether the GET 2020 goals had been met in P . R . China , we conducted a targeted assessment with national scope . Province assessment teams , trained in WHO Trachoma Rapid Assessment ( TRA ) methodology and in WHO simplified trachoma grading system , carried out assessments in 16 provinces ( among them , 2 provinces conducted pilot assessment ) . Based on the published literature , including national and international reports , suspected trachoma-endemic areas within each province were identified . Within these areas , trachomatous inflammation- follicular ( TF ) assessments were carried out in at least 50 grade-one children in primary schools serving villages with the lowest socio-economic development . Trachomatous trichiasis ( TT ) and corneal opacity ( CO ) assessments were conducted among persons aged 15 and over in villages within the catchment area of the selected schools . Of 8 , 259 children examined in 128 primary schools in 97 suspected trachoma endemic areas , only 16 cases of conjunctivitis were graded as TF . 38 cases with TT were found among the 339 , 013 examined residents in villages surrounding the schools . Among these 97 suspected trachoma endemic areas in only three was the prevalence of TT more than 0 . 2% . This large study suggested that trachoma was not a public health problem in 16 provinces that had been previously suspected to be endemic . These findings will facilitate planning for elimination of trachoma from PR China .
China used to be among the countries with a high prevalence of trachoma . Since the new China was founded in 1949 , the government has actively supported trachoma control activities . In 1960s , the Chinese government adopted a policy to strengthen healthcare provision in rural areas , which in turn provided an opportunity to increase trachoma control efforts in rural areas . Several epidemiological surveys in the 1990’s showed that the prevalence of trachoma had been drastically decreased from earlier years and its severity reduced in terms of trachomatous trichiasis cases reported in medical records . According to those surveys , trachoma prevalence among primary and secondary school students was 16% and 18% , respectively , in 1992; dropping to 11% and 14% in 1995 , and to 8% and 8% in 2000 [1] . In 1999 at the invitation of the Chinese Ministry of Health , the World Health Organization ( WHO ) organized a national workshop on trachoma control to review the status of the disease and accelerate the elimination of blindness from trachoma in China [2] . Since then , prevention of blindness activities in China was implemented by the Chinese government and international partners ( e . g . The Lions Clubs International ) . Surgical service for cataracts and elimination of trachoma were among the priorities . The number of trachoma cases identified in public health activities , in surveys on causes of visual impairment and in clinical care services were steadily decreasing–consistent with a rapid socio-economic development , significant improvement of personal hygiene ( particularly in schools ) , as well as increased access to eye care service across China . The China Nine-Province Survey in 2006 [3 , 4] , the Second National Survey on Disabled Persons in 2006 and other data provided evidence that trachoma was no longer a common cause of vision loss [5 , 6] , despite some articles reporting the finding of active trachoma cases in schools [7 , 8] . At the launch of the Global Elimination of Trachoma ( GET ) 2020 campaign by the WHO in 1997 , China was included on the list of the countries to be verified for possible presence of trachoma based on historical epidemiological data [9] . Although data from blindness surveys and routinely collected clinical data were showing that blindness from trachoma was no longer a public health issue , it was decided to assess the current situation using internationally adopted epidemiological tools , definitions and standards , including the WHO Simplified Trachoma Grading System [10] . The thresholds for elimination of trachoma as a public health problem include: i . e . 1 ) TF prevalence of <5% in children 1–9 years; 2 ) prevalence of TT in people aged 15 years or more of <0 . 2%; and 3 ) evidence that the health system is able to identify and manage incident TT cases [11] . Despite current recommendations on population-based prevalence surveys , TRA was used in this survey in view of the enormity of such a task in the most populous country China . In 2012 , the former Chinese National Health and Family Planning Commission , with WHO technical support and funding from Lions Clubs International , undertook the assessment of the elimination of trachoma in P . R . China . By the end of the year 2015 , the assessment was completed and this article reports main results .
TRA was conducted in all provinces historically known , or suspected of having trachoma-endemic areas . Identification of the provinces to be surveyed was based on two approaches: a ) Identification based on published historical data and reports on trachoma . Our study fully considered those existing studies and papers , but the publication dates of those were different , and survey strategy and definition standards used were substantially different ( e . g . the Chinese trachoma grading system vs . the WHO grading system ) . In our study , with the support of WHO experts , we adopted for the first time the WHO international trachoma grading standard . Regarding our selection of survey location , we started from the Report of the first meeting of the WHO alliance for the global elimination of trachoma published in 1997 . In this workshop , representatives from all provinces in China reported the historic and current information about epidemic condition of trachoma . Based on that report [2] , there were 12 provinces with suspected trachoma endemic areas–Hebei , Inner Mongolia , Liaoning , Anhui , Henan , Hainan , Guizhou , Yunnan , Shaanxi , Gansu , Qinghai and Ningxia . b ) Identification based solely on low socio-economic status and access to water and sanitation . This led to the addition of 2 autonomous regions ( organizationally equivalent to provinces ) —Guangxi and Tibet—based on ranking among the 5 provinces with the lowest gross domestic product ( GDP ) and per capita income in 2013 . Additionally , 2 provinces—Shandong and Sichuan—were included as pilot assessment provinces to examine our method of study . After the pilot test , our methodology was improved and perfected . TRA was also conducted in these pilot assessment provinces , however the results should be presented separately . Each province was divided into several geographic areas with a population of approximately 100 , 000 to 150 , 000 , using county and township boundaries and census data . Suspected trachoma endemic areas within these geographic population clusters were determined in the following ways: a ) Evidence regarding the presence of trachoma based on the published literature; b ) Existing data on trachoma among primary-school students over the past five years: c ) Existing routinely-collected data on trachoma found in patients presenting to eye care services . TRA was carried out in selected primary schools in the identified suspected endemic areas within each of the selected provinces and included: a ) An assessment for TF in at least 50 grade-one school children within the surveyed areas . B ) An assessment of TT and CO in persons aged 15 years or over in villages within the catchment area of the surveyed schools . The WHO simplified Trachoma Grading System [10] was used to provide consistency with internationally agreed approaches for trachoma assessment and with criteria set by the WHO Alliance for the Global Elimination of trachoma by 2020 [12] . Case finding focused on three grades of trachoma: TF , five or more follicles more than or equal to 0·5mm on the upper tarsal conjunctiva; TT , at least one ingrown eyelash touching the globe due to trachoma , or evidence of epilation; CO , corneal opacity blurring part of the pupil margin due to trachoma [13] . In each selected province , 2 or 3 provincial survey teams were assembled , equipped and trained . Each provincial team included 3 certified ophthalmologists , who were ultimately responsible for trachoma grading in the assessment . Each ophthalmologist was equipped with a head-mounted magnifier ( ×2·5 ) , torch , and alcohol-based hand gel for sanitizing hands . Ophthalmologists of each survey team attended a national two-day structured training course in Beijing first , then a two-day field training in a county in Inner Mongolia autonomous region , where more TF cases may existed according to local ophthalmologist’s opinion . The national-level trainers included a WHO officer and international trachoma experts . Training included the WHO SAFE strategy , the WHO TRA methodology , the WHO Simplified Trachoma Grading System , laboratory confirmation of Chlamydia trachomatis as well as procedures for identification of population to be surveyed . This training included tests-of-agreement for trachoma identification and grading between each trainee and the WHO trachoma experts . All trainees entered into 2 classrooms and examined 80 schoolchildren , who were healthy , had conjunctivitis or were real trachoma cases ( only 2 positive TF cases existed in each classroom , since we hardly to find TF cases . In fact , we collected TF cases in several primary schools in Inner Mongolia together for this agreement-test ) . Each team member was observed by a WHO expert grader who assessed agreement . All 54 trainees passed the agreement testing as they were measured totally equivalent to the WHO expert grader . After academic and field training at the national level , two-day provincial-level training was organized in each province to educate local ophthalmologists and supporting staff regarding TRA process [13] . Ophthalmologists who attend the national training course were the trainers . However , only the ophthalmologists who attend the national training course are responsible to perform the examination , trachoma diagnosis and grading in the TRA assessment . The purpose of provincial training was to equip local ophthalmologists and supporting staff professional and standardized method in the TRA assessment . The national and provincial training course included the detection of TT and CO . The diagnostic criteria of TT and CO are relatively easy for ophthalmologists . However , trichiasis may be caused by other situations , such as senile or spasmodic trichiasis , CO may be caused by viral or bacterial infection . Thus , we emphasize to do differential diagnosis in training and in the assessment . If the person is absence of TS , it may indicate that the trichiasis or CO was not trachomatous in origin . Since in the training practice in Inner Mongolia , there are too few cases of TT and CO , thus , we didn’t perform TT and CO agreement test . The trained provincial survey teams began field work by visiting ophthalmologists in county or local leading hospitals within the suspected endemic areas to discuss their trachoma knowledge and enquire about TF diagnoses or TT surgeries performed in recent years . Hospital’s records were also checked regarding diagnosis of corneal blindness due to trachoma . The TRA was carried out in the identified suspected endemic areas within each of the selected provinces and included: a ) An assessment for TF in at least 50 grade-one school children within the surveyed areas . B ) An assessment of TT and CO in persons aged 15 years or over in villages within the catchment area of the surveyed schools . Even in the complete absence of evidence or local awareness regarding trachoma cases , one or more primary schools with the worst socio-economic conditions , worse water supply and poorest schooling conditions were selected for TRA in each of the selected areas . Priority was given to boarding schools with a high enrollment of children from rural families as trachoma is easily transmitted in centralized environment . There was no gender bias as quantities of male students and female students in each school were very close . At each selected school , survey team was mandated to examine at least 50 grade-one students ( usually seven-year-old ) . School teachers were fully mobilized to encourage students and their family members to participate to address potential sources of bias . In order to reduce the risk of missing possible endemic areas , two alternative strategies shown in Fig 1 were used in assessing the presence and detailed number of TF in the selected schools , depending on whether TT or CO were found in the desk review of hospital records within the suspected endemic area . The TRA approach was used when no records of TT or CO were found in any of the hospital reviews . The second strategy was used in schools when one or more records of TT or CO were found in any of the surveyed hospitals . TRA for TT and CO was conducted in two ways: a ) If recorded cases of TT or CO were found during the review of patient records in any of county/lead hospitals within suspected endemic areas , these individuals were contacted for reexamination/confirmation and the affected village ( s ) was screened for additional cases of TT or CO among residents ( defined as people living in the village for the last six months ) aged 15 years or older; b ) If TT or CO cases were not found during hospital records reviews , survey team visited at least one village surrounding the school assessed for TF to look for TT or CO among residents aged 15 years or older . In both assessment scenarios , known or suspected TT or CO cases were identified by consulting village leaders and doctors , followed by a visit to the home of the TT or CO suspect to confirm the diagnosis . Village leaders and doctors after standardized training were fully mobilized to participate as assistants for the assessment work in the village and inspection rate was guaranteed to address potential selection bias . The village leaders and doctors were familiar with every family and residents in the village , since the population of villages is usually less than 1 , 000 . After the assessment for TT and CO in suspected endemic areas , an approximate prevalence rate for TT and CO in the area was calculated as the number of confirmed cases divided by the estimated number of persons aged 15 or above in the village . As data from each suspected areas became available , the national project office analyzed the findings to decide whether a population-based survey was needed . According to the project plan , a population-based survey was considered warranted if more than 4 TF cases were identified among the 50 children examined at any one school . Because there was no school with more than 2 TF cases , a population-based survey for TF was not undertaken in any of the suspect endemic areas . If an area was found to have more TT or CO cases , we should expand survey area . If the prevalence ( unadjusted by sex and age ) of TT or CO is higher than the threshold 0 . 2% , the assessment was also conducted in nearby areas . If the prevalence of TT or CO was confirmed more than the threshold 0 . 2% , then this area needs to conduct the population-based survey for TT or CO . After TRA , we just found a few TT cases . For further confirming the situation of TT and getting more information of TT , we mobilized local ophthalmologists who attended provincial training to conduct a population-based survey for evaluating the situation of TT more broadly . For TF-positive children a provision of drugs for treatment was secured through county resources . The treatment regimen was the one recommended by WHO GET2020 campaign–one-gram of oral azithromycin , either tablet or suspension , for adults and 20 mg per kg body weight for children . Family members of TF patient were also to be treated . For TT cases , a bilamellar tarsal rotation procedure or other similar oculoplastic method was provided free of charge to the patient . Electrolysis was used for the patients with one or two inverted eyelashes touching the globe ( not the cornea ) in the absence of entropion . County hospitals were selected to provide the required surgery . Strict quality control measures were undertaken: a ) The project design was based on WHO standards for trachoma control ( TRA and Simplified Grading System ) . b ) A pilot study was undertaken to verify the feasibility of the planned assessment methodology . c ) National and provincial training sessions were held for survey personnel with formal inter-observer agreement testing of grading consistency . d ) Supervisory activities were implemented throughout the field assessment process with monitoring and participatory supervision by national and international experts . Each provincial survey team had a dedicated analyst responsible for data recording , computerized data entry , and data review for completeness and accuracy . EpiData 3 . 0 was used for data entry and analysis . This study basically adopted routine public health measures . All involved in the intervention were provided with written informed consent which asked participants themselves or juveniles’ parents to sign on it . IRB of Peking Union Medical College Hospital approved the ethical review of elimination of blinding trachoma by 2016 in China on Feb . 6 , 2013 . The statement shown in the certificate is “this project is designed scientifically , risks and benefits of human subjects are rational , and letter for informed consent meets the ethical requirement” . All data analyzed were anonymized . http://dx . doi . org/10 . 17504/protocols . io . qu9dwz6
From June to September 2013 , Shandong and Sichuan provinces were selected to conduct pilot assessment . Table 1 showed the information from the pilot assessment . 7 suspected areas were identified in Shandong provinces , and 887 students aged about 7 from 14 primary schools were examined for TF . No TF cases were found . 18 , 180 villagers aged 15 or above from 25 villages nearby the selected primary school were examined for TT or CO . Only 1 TT case and no CO case were found . 7 suspected areas were identified in Sichuan province , and 350 students aged about 7 from 7 primary schools were examined for TF . No TF cases were found . 3 , 337 villagers aged 15 or above from 12 villages nears the selected primary school were examined for TT or CO . 1 CO case was found , however , 14 TT cases among 1005 persons in one area and 2 TT cases among 116 persons in one area were found , the prevalence was 1 . 393% and 1 . 724% respectively . The prevalence was unadjusted for age and gender . A total of 83 suspected trachoma-endemic areas were identified in those 14 provinces in mainland China . Using TRA methodology at least 50 grade-one children ( generally seven years of age ) in each of 107 schools were assessed for TF . 7 , 022 children received the assessment , only 16 cases of TF were found in 14 schools , in which TF cases were found in only 2 schools were 2 cases found; in all of the other schools had only 1 case . ( Table 2 ) No TF cases were found in 93 schools among 107 schools , . The threshold of four TF-positive children for further epidemiological investigation and PCR testing for Chlamydia trachomatis was found nowhere . Within the 83 suspected trachoma-endemic areas , TRA for TT and CO was carried out in 161 villages in 14 provinces with 317 , 496 examined persons aged 15 years or older . In Tibet , there was no villages near two primary schools , thus the TRA for TT and CO cannot performed there . A total of 21 cases of confirmed TT , who were all 50 years old or above were found in 5 provinces ( Inner Mongolia , Anhui , Guizhou , Ningxia , Guangxi ) . Only in 3 areas ( Danshan Town and Ningnan County in Sichuan , and Baochang Town in Inner Mongolia ) , the prevalence of TT was above the threshold 0 . 2% ( Table 2 ) . Across all of the 161 villages , only 1 case of CO was found , the prevalence of CO is less than 0 . 1% . After TRA , we mobilized local ophthalmologists to conduct population-based survey for TT in the provinces with TRA and pilot study . These local ophthalmologists got a training course about diagnosis of TT and population-based survey . Since there were no enough local ophthalmologists in Tibet and Guangxi to conduct this survey , so this survey was conducted in only 14 provinces . Table 2 showed the results . We found 1 , 296 TT cases in 87 , 540 , 342 persons aged 15 years or above in 55 , 481 villages . Although the prevalence of TT was above 0 . 2% in three TRA areas ( Tables 1 and 2 ) , the province-level prevalence of TT was <0 . 2% across all the 16 provinces ( Table 3 ) . The prevalence was unadjusted for age and gender . Therapeutic interventions for TF and TT patients in the 16 provinces were completed by December 2014 . The 16 TF patients were cured with azithromycin . 1 , 334 TT cases ( including 17 cases in pilot assessment , 21 cases in TRA , and 1296 cases in the population-based survey by local ophthalmologists ) received successful treatment .
WHO document in 2012 stated that trachoma was still considered prevalent in 8 countries , among the 27 countries in the Western Pacific region , including China . Based on reports for trachoma control efforts and relevant statistics in China during recent decade , it was on the contrary understood that trachoma had been effectively controlled and eliminated as a cause of vision loss of public health importance . There was no record of trachoma or TT in the last decade in medical reports from county and provincial hospitals , despite a quite exhaustive health data reporting system . In the absence of an assessment based on WHO recommended methodologies , it was understandable that China was included among those countries suspected of having trachoma as a public health problem . Furthermore , a new trachoma survey was a good opportunity to introduce the standardized WHO Simplified Grading System , which was not commonly used in China , leading to inaccurate reporting of the disease against the internationally adopted criteria . A trachoma assessment would provide the basis upon which to determine if any further action was required to demonstrate that the goal of WHA Resolution 51 . 11 had been reached . China is a large geographic area with a huge population and substantial regional differences in socio-economic and hygienic conditions . A project of this scope to establish the public health priority that should be assigned to the elimination of trachoma under such a varied geographical and demographic environment was unprecedented . Although TRA has never been validated for the assessment of elimination of trachoma as a public health problem , and because of the absence of any trachoma control intervention for many years , it was considered the most suitable approach to identify remaining endemic areas of trachoma . In each suspected trachoma-endemic area , one or more primary schools with the worst socio-economic conditions , worse water supply and poorest schooling conditions were therefore selected for TRA . TRA findings were then used to determine whether a standard population-based survey was needed to confirm the existence and severity of trachoma . This two-step strategy was the key in the design of a feasible approach to establishing whether trachoma was still a public health issue in China . With a threshold of 4 cases of TF among 50 children examined in any school to start a population-based survey , none of the 97 suspected areas warranted such a survey . Our results show that nowhere near the location with the number of TF cases required for public health concern were found in any of the suspected areas . Although we found three areas ( two areas in the pilot assessment , one areas in TRA ) , which belong in the remote and poor areas , with the prevalence of TT higher than 0 . 2% , after examining other areas nearby , there were fewer TT cases and the result didn’t reach our threshold to conduct the population-based survey for TT . After TRA , we mobilized local ophthalmologists to conduct population-based survey for TT in the provinces with TRA and pilot study , and found that the prevalence of TT in each surveyed province was not more than 0 . 2% . Moreover , trachoma was not a cause of visual impairment based on the results of national epidemiological surveys and rapid assessment for avoidable blindness . And eye care system at each level was able to manage existing and incident cases . So no further investigation was carried out , but more attention would still be given to these areas in the future work , especially the remote and poor villages . With negative historical record relevant to trachoma in the other 15 provinces of China and with socio-economic development , improvement of hygiene and GDP ranking better than in the 16 surveyed provinces , it may be concluded that trachoma is no longer a public health issue in China . There are some limitations in our paper , and the main one is the coverage area and population in the assessment . Based on our study design , we conduct TRA in which just those selected districts and people were examined . And the conclusion we reach is based on TRA rather than population-based survey . Despite the positive result from the assessment , proper screening and improvement in hygiene condition should still be maintained , and continuous monitoring on trachoma epidemic condition would also be conducted . Our methodology is generalizable among countries with large population , vast territory and great regional diversity which can use TRA for initial assessment and then population-based survey when needed . There are several underlying reasons that support our findings of trachoma elimination in China . First , China has enjoyed rapid socio-economic development during the recent decades with significant improvement in income and living conditions , including quality of housing and water supply . From 1990 to 2014 , population benefited from water improvement programs of varied types were increased by 37% , with the number of nearly 249 million persons . And 262 million households in rural areas were benefited from latrine improvement programs . Second , with an expanding public health system , the Chinese government has attached substantial importance to implementation of basic hygiene measures , which are also effective for the elimination of trachoma . Third , the centers for disease prevention and control , maternal and child care centers and health education organizations have been active at all levels in working with professional institutions in trachoma control . Fourth , there have been consistent efforts in trachoma control by the Chinese Society of Ophthalmology . Currently , ophthalmologists in China actively participate into a variety of activities for blindness prevention , including frequent screening for eye diseases , free treatment for TF patients by Azithromycin , health education and publicity at community level . Fifth , hygiene in rural environments has improved in conjunction with public works directed at drinking water and sanitation improvements . Sixth , the massive use of antibiotics is also likely to have curbed the transmission of Chlamydia . Last , but not least , results from 16 Rapid Assessments of Avoidable Blindness and recent epidemiological surveys also provided strong evidence that trachoma is not a significant cause of blindness [4 , 5 , 15 , 16] . This large study suggested that trachoma was not a public health problem in 16 provinces that were previously suspected to be endemic . These findings will facilitate planning for elimination of trachoma from P . R . China . | China used to a country with a high burden of trachoma , but recent empirical observation and routinely collected eye care data were suggesting that elimination of trachoma had been achieved in China . In order to verify that and reach the target set in the WHA 51 . 11 Resolution , we have assessed current situation using the World Health Organization’s Simplified Trachoma Grading System and TRA . As China is a country with a big population and huge regional disparity , this national condition is also considered in methodology design and implement . Among 8 , 259 children examined in 128 primary schools in 97 suspected trachoma endemic areas , only 16 cases of conjunctivitis were graded as TF . 38 cases with TT were found among the 339 , 013 examined residents in villages surrounding the schools , among these 97 suspected trachoma endemic areas only in 3 areas , the prevalence of TT was more than 0 . 2% . Trachoma is no longer a public health issue in China , however , we still need to pay attention to TT in the remote and poor areas . Therapeutic interventions for few residual patients have been provided for free , and health education and publicity on trachoma prevention have been continued in the whole society . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"standardized",
"tests",
"medicine",
"and",
"health",
"sciences",
"education",
"china",
"sociology",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"bacterial",
"diseases",
"eye",
"diseases",
"global",
"health",
"neglected",
"tropical",
"dise... | 2019 | Assessment of trachoma in suspected endemic areas within 16 provinces in mainland China |
Pseudomonas aeruginosa , an important opportunistic pathogen of man , exploits numerous factors for initial attachment to the host , an event required to establish bacterial infection . In this paper , we rigorously explore the role of two major bacterial adhesins , type IV pili ( Tfp ) and flagella , in bacterial adherence to distinct host receptors at the apical ( AP ) and basolateral ( BL ) surfaces of polarized lung epithelial cells and induction of subsequent host signaling and pathogenic events . Using an isogenic mutant of P . aeruginosa that lacks flagella or utilizing beads coated with purified Tfp , we establish that Tfp are necessary and sufficient for maximal binding to host N-glycans at the AP surface of polarized epithelium . In contrast , experiments utilizing a P . aeruginosa isogenic mutant that lacks Tfp or using beads coated with purified flagella demonstrate that flagella are necessary and sufficient for maximal binding to heparan sulfate ( HS ) chains of heparan sulfate proteoglycans ( HSPGs ) at the BL surface of polarized epithelium . Using two different cell-free systems , we demonstrate that Tfp-coated beads show highest binding affinity to complex N-glycan chains coated onto plastic plates and preferentially aggregate with beads coated with N-glycans , but not with single sugars or HS . In contrast , flagella-coated beads bind to or aggregate preferentially with HS or HSPGs , but demonstrate little binding to N-glycans . We further show that Tfp-mediated binding to host N-glycans results in activation of phosphatidylinositol 3-kinase ( PI3K ) /Akt pathway and bacterial entry at the AP surface . At the BL surface , flagella-mediated binding to HS activates the epidermal growth factor receptor ( EGFR ) , adaptor protein Shc , and PI3K/Akt , and induces bacterial entry . Remarkably , flagella-coated beads alone can activate EGFR and Shc . Together , this work provides new insights into the intricate interactions between P . aeruginosa and lung epithelium that may be potentially useful in the development of novel treatments for P . aeruginosa infections .
Pseudomonas aeruginosa is an opportunistic human pathogen associated with a broad spectrum of life-threatening infections in the setting of epithelial injury and immunocompromise ( reviewed in [1] ) . This gram-negative pathogen ranks among the leading causes of hospital-acquired pneumonia , urinary tract infections , bloodstream infections , and surgical site infections . In addition to their frequent occurrence , nosocomial P . aeruginosa infections are often severe , with an excess attributable mortality rate of almost 50% for mechanically ventilated patients with P . aeruginosa pneumonia [2] . The bacterium is also the leading cause of respiratory morbidity and mortality in patients with cystic fibrosis ( CF ) [3] , [4] , as well as a frequent cause of exacerbations in individuals with advanced chronic obstructive pulmonary disease [5] . P . aeruginosa infections are also reported as a complication of HIV infections and are becoming more frequent as patients with AIDS survive longer [6]–[8] . Notably , therapeutic options are becoming increasingly limited with the continued emergence and spread of multi-drug resistant strains . Thus , increasing our understanding of the pathogenesis of P . aeruginosa infections is critical for the development of new therapeutics that target this medically important pathogen . The pathogenesis of P . aeruginosa infections is multifactorial and complex . Bacterial attachment is an initial and critical step that involves complex interactions between bacterial adhesins and host receptors either on the apical ( AP ) or basolateral ( BL ) surface of polarized epithelium . Using cultured epithelial cells grown as polarized monolayers , which recapitulate simple epithelial tissue , or as three-dimensional cysts , which mimic the organization of simple epithelial organs , we have recently demonstrated that N-glycans are necessary and sufficient for bacterial binding and consequent entry and cytotoxicity at the AP surface of polarized epithelium [9] . In contrast , heparan sulfate ( HS ) chains of heparan sulfate proteoglycans ( HSPGs ) are necessary and sufficient to mediate these events at the BL surface of polarized cells . We showed that in incompletely polarized cells , a model for tissue injury , HSPGs are upregulated at the AP surface , which leads to enhanced binding and subsequent tissue damage by P . aeruginosa . These results provide an explanation , at least to some extent , for the increased susceptibility of injured tissue to P . aeruginosa infections . Although our previous work characterized distinct AP and BL host receptors , the downstream host signaling pathways associated with these critical events and bacterial binding partners remained to be elucidated . Two major adhesins have been identified in P . aeruginosa , flagella and type IV pili ( Tfp ) [10]–[12] . The single polar flagellum is a polymer composed of flagellin , the product of the fliC gene , although a large number of gene products are required for flagellar assembly and function [13] . Flagella are required for adhesion to cells , swimming motility , and biofilm formation . In addition , monomeric flagellin is recognized by the innate immune system , either by binding to Toll-like receptor 5 ( TLR5 ) at the cell surface or by recognition of individual subunits by intracellular cytosolic sensors [14]–[18] Mechanistic or structural details of the interaction of flagella with the host epithelium are still lacking . The flagellar cap protein of P . aeruginosa strain O1 ( PAO1 ) , but not other strains , binds to LewisX oligosaccharides in mucins [19] , but whether this is relevant to binding to host epithelial cells is unknown . Tfp are polarly localized appendages composed of pilin polymers that undergo reversible assembly and disassembly , allowing the bacteria to move over a solid surface in a process termed twitching motility ( reviewed in [20] ) . Tfp also function as phage receptors , contribute to early steps in biofilm formation , and serve as adhesins to mammalian cells [21] . Several studies have identified different glycosphingolipids as host receptors for Tfp-mediated binding at the AP surface of polarized cells [22] , [23] , although their roles in mediating bacterial binding remain controversial [24] , [25] . Following adhesion to host epithelium , P . aeruginosa can induce host cell death or enter non-phagocytic cells ( reviewed in [10] , [26] ) . Internalization may permit the bacteria to penetrate the epithelial cell layer , reach the bloodstream , and disseminate to distant organs and/or it may represent a host defense mechanism that contributes to bacterial clearance [10] , [27] . The molecular events underlying P . aeruginosa invasion into non-phagocytic cells are incompletely understood . P . aeruginosa entry is an actin-dependent process that involves Rho family GTPases [28] , activation of tyrosine kinases , such as Src [29] , [30] or Abl [31] kinases and subsequent tyrosine phosphorylation of several host proteins , including caveolin [32] . We have previously shown that phosphatidylinositol 3-kinase ( PI3K ) and its effector protein Akt ( also known as the serine threonine protein kinase B ) are necessary and sufficient for and are activated upon bacterial internalization into Madin Darby Canine Kidney ( MDCK ) cells [33] . However , specific AP or BL upstream receptors associated with this event have not yet been identified , and the PI3K/Akt pathway can be activated by many stimuli , including growth factor receptors , such as the epidermal growth factor receptor ( EGFR ) [34] . Furthermore , the role of bacterial ligands , e . g . Tfp or flagella , in these newly described signaling events has also not been investigated . In this work , we characterize important bacterial and host factors that play an essential role in the complex binding and signaling networks utilized by P . aeruginosa . We rigorously assess the role of Tfp and flagella in mediating bacterial binding to specific host N-glycans and HSPGs at the AP and BL surfaces , and test whether such interactions dictate activation of specific signaling pathways . We demonstrate that Tfp are necessary and sufficient to mediate maximal bacterial binding to N-glycans at the AP surface , while flagella are necessary and sufficient to mediate maximal bacterial binding to HS chains of HSPGs at the BL surface of polarized airway epithelium . We find that P . aeruginosa internalization at the AP surface is dependent on Tfp binding to N-glycans and , to some extent , on activation of PI3K and Akt . P . aeruginosa internalization at the BL surface is dependent on flagella binding to HS followed by activation of EGFR and PI3K/Akt pathway . Remarkably , flagella-coated beads alone are sufficient to trigger EGFR phosphorylation and activation of downstream adaptor protein . Our work identifies key factors and interactions required for establishing P . aeruginosa attachment and internalization , affording new avenues for development of treatments for acute and chronic P . aeruginosa infections .
Our previous studies established that P . aeruginosa binds preferentially to N-glycan chains at the AP surface of polarized epithelium , with preferential affinity for more complex chains after up-regulations of N-glycosylation [9] . At the BL surface , the bacterium binds preferably to HS chains of HSPGs . To determine whether flagella or Tfp , the major P . aeruginosa adhesins , differentially mediate binding to these distinct AP and BL host receptors , we utilized isogenic mutants of PAO1 in which the gene encoding PilA , the major subunit of the Tfp ( PAO1ΔpilA ) or the gene encoding FliC , the major subunit of flagella ( PAO1ΔfliC ) , is deleted . Standard adhesion assays , in which bacteria were added for 2 h to the AP or BL surface of polarized Calu-3 cells grown as polarized monolayers on Transwell filters , were performed [9] . While other P . aeruginosa adhesins have been identified , such as the cup fimbrial adhesins [35] and lectins PA-IL ( LecA ) and PA-IIL ( LecB ) [36] , Tfp and flagella were the predominant adhesins under the conditions of our experiments , as the PAO1ΔfliCΔpilA double mutant exhibited undetectable binding ( data not shown and [21] ) . Consistent with our previously published studies utilizing PAK [9] , PAO1 bound approximately 2-fold more efficiently to the BL surface than to the AP surface of polarized airway epithelium ( Figure 1A ) . Both the Tfp mutant ( PAO1ΔpilA ) and the flagella mutant ( PAO1ΔfliC ) bound less efficiently to the AP or BL surfaces of the epithelium when compared to PAO1 ( Figures 1A–C ) , suggesting that the absence of either of these two adhesins impacts bacterial binding and that Tfp and flagella may function synergistically . Importantly , PAO1ΔfliC still bound almost 2-fold better than PAO1ΔpilA to the AP surface ( P<0 . 05 ) , but it bound ∼9-fold less efficiently than PAO1ΔpilA to the BL surface ( P<0 . 05 ) ( compare Figures 1B and C ) . Furthermore , while PAO1ΔpilA bound ∼3-fold more efficiently to the BL surface than to the AP surface of polarized epithelium ( Figure 1B ) , PAO1ΔfliC adhered over 5-fold better to the AP surface than to the BL surface ( Figure 1C ) . These results suggest that Tfp are the predominant adhesin at the AP surface whereas flagella function as the major adhesin at the BL surface . We then investigated the role of host N-glycans in Tfp- or flagella-mediated binding . N-glycans are found on both AP and BL surfaces of polarized epithelium whereas HS chains are expressed predominantly on the BL surface of polarized epithelium [9] . When the expression of more complex N-glycans was triggered by long-term culture of Calu-3 cells in the presence of mannose ( Man ) [37] , we observed a 2-fold increase in the binding of PAO1 or PAO1ΔfliC to the AP surface of polarized airway epithelium . However , no effect on the BL binding of PAO1 or PAO1ΔfliC was observed ( Figure 1C ) . In control experiments , long-term culture of Calu-3 cells in the presence of galactose ( Gal ) , which does not enhance N-glycosylation , had no effect on bacterial adhesion ( data not shown ) . Inhibition of N-glycosylation by pre-treatment of Calu-3 cells with tunicamycin , which decreased N-glycosylation by 50% under the conditions of our experiments ( more extensive deglycosylation disrupted the monolayer integrity ) , decreased the AP adhesion of PAO1 ( Figure 1A ) and PAO1ΔfliC ( Figure 1C ) by 2 . 5-fold , but had no effect on BL binding . These treatments did not have statistically significant effects on binding of PAO1ΔpilA to the AP surface of polarized cells ( Figure 1B ) . Together , these results suggest that Tfp bind primarily to N-glycans at the AP surface of polarized epithelial cells . We used two approaches to determine the role of host HS chains of HSPGs in Tfp- and flagella-mediated bacterial adherence . First , addition of excess heparin competitively inhibited the binding of PAO1 and PAO1ΔpilA at the BL surface of Calu-3 cells ( Figures 1A and B ) , but had no effect on binding of PAO1ΔfliC to the BL surface ( Figure 1C ) . Since HSPGs are predominantly expressed on the BL surface , it was not surprising to observe that exogenous addition of heparin had little effect on binding of PAO1 or of either adhesin mutant to the AP surface of polarized cells . To rule out non-specific charge effects , we demonstrated that addition of another highly negatively charged glycosaminoglycan chain , chondroitin sulfate ( CS ) , had no effect on binding of PAO1 , PAO1ΔpilA , or PAO1ΔfliC to either surface ( data not shown ) . Second , pre-treatment of cells with heparinase III , an enzyme that cleaves HS chains , had a similar effect on bacterial adhesion as did addition of excess heparin . It reduced adhesion of PAO1 and PAO1ΔpilA to the BL surface of polarized Calu-3 cells but had no effect on adhesion of PAO1ΔfliC . Heparinase III treatment had no significant effect on the binding of any of the strains to the AP surface of polarized Calu-3 cells . Enzymatic removal of CS by chondroitinase ABC did not alter bacterial attachment at either surface ( data not shown ) , confirming the specific role of HS in flagella-mediated binding of P . aeruginosa . Together , these results demonstrate P . aeruginosa binding at the BL surface is predominantly mediated by flagella interactions with HS chains of HSPGs . Two different in vitro binding affinity assays were utilized to rigorously determine whether Tfp or flagella were sufficient to mediate binding to N-glycans or HS , respectively . We have previously developed a fluorometric assay that quantifies bacterial attachment to plastic wells coated with various glycans [9] . Our studies revealed that GFP-expressing PAO1 binds in a dose-dependent manner to plastic wells coated with HS or to a complex hybrid N-glycan chain ( ( Gal-GlcN ) 4Man3 ( GlcN ) 2 ) , with the strongest binding observed to HS . Here , we modified this assay; in place of bacteria , we used 2-µm green fluorescent beads coated with purified flagella or Tfp , isolated by shearing from PAO1ΔpilA or PAO1ΔfliC , respectively ( Figure S1 ) . Coommasie Blue staining of the purified adhesin preparations electrophoresed on SDS-PAGE did not reveal any contaminating products ( Figure S1 ) . We note that the sheared adhesins may be composed of short polymers ( i . e . flagella or Tfp ) or the individual subunits ( flagellin or pilin ) . For ease of clarity , we will refer to them as flagella or Tfp . As shown in Figure 2B , flagella-coated beads bound strongly in a dose-dependent manner to HSPG or HS chains alone and at low levels to different N-glycans ( structures shown in Figure 2A ) , but not measurably to single sugars ( Man , N-acetylated glucosamine ( GlcNAc ) , fucose Fuc , or galactose ( Gal ) ) , CS , or to non-sulfated hyaluronic acid ( HA ) . These results suggest that i ) flagella are capable of strong binding to HS chains , that ii ) sulfate groups of HS provide binding sites and/or the anionic charge of sulfate is necessary for the interaction , and that iii ) flagella may also bind to sugar sequences along the N-glycan chain in a polyvalent manner . We next examined the binding specificity of beads coated with sheared Tfp . In contrast to the results obtained with flagella-coated beads , Tfp-coated beads bound most avidly to N-glycan chains ( Figure 2C ) , with the strongest binding to N-glycan-3 , the most complex N-glycan chain ( Figure 2A ) . There was minimal binding to the single sugars ( Man , GlcNAc , Fuc , or Gal ) indicating that single sugars are not sufficient to mediate binding to Tfp . Likewise , there was almost no binding to HSPGs , HS , or other glycosaminoglycans , suggesting that Tfp almost exclusively recognize N-glycan chains . Identical results were obtained with beads coated with flagella or Tfp isolated from strain PAK ( Figure S2 ) and with pili isolated from strain PA103 ( Figure S3 ) , suggesting that the observed binding specificities are a general property of the adhesins and are not strain-specific . As a further test for the specificity of the binding of Tfp-coated beads , we tested whether the C-terminus of Tfp , which contains the binding determinants , was required . For these experiments , we isolated sheared surface Tfp from a PA103 mutant , which harbors a transposon insertion 13 amino acids from the C-terminus ( PA103 Mutant 9 [38] ) ( Figure S3 ) . The mutant pilin is predicted to be truncated between the two C-terminal cysteines required for pilin folding [39] . Beads coated with Tfp isolated from PA103 pili ( Figure S3 ) demonstrated a similar binding specificity as beads coated with Tfp from PAO1 ( Figure 2 ) or PAK ( Figure S2 ) . In contrast , the Mutant 9 Tfp-coated beads bound poorly to N-glycans ( Figure S3 ) . These results suggest that three-dimensional structure of pili is not compromised under the conditions of our experiments and that the C-terminal binding determinants are required for in vitro binding of Tfp to N-glycans . We extended our studies using a complementary in vitro assay that measures bead aggregation under defined shear forces to detect molecular interactions [40] . As shown in Figure 2D , we first tested the interaction of flagella-coated green fluorescent beads with red fluorescent beads coated with various glycans . In control experiments in which equal numbers of flagella-coated green beads were mixed with BSA-coated red beads , 99% of the aggregates comprised a single color ( 52% green and 47% red ) and only 1% were yellow . These results indicate that there is minimal non-specific bead aggregation under the conditions of our experiments . In contrast , when flagella-coated green beads were mixed with HSPG-coated red beads , 80% of the aggregates were yellow . This interaction was inhibited by exogenous addition of heparin , with the number of yellow aggregates decreasing to ∼10% . Aggregation of flagella-coated green beads with N-glycan-3-coated red beads resulted in 34% yellow aggregates , which was decreased to 16% when excess N-glycan-3 was added to competitively inhibit binding . Together , both of these in vitro assays confirm our cell-culture based experiments and demonstrate conclusively that flagella can bind directly to HS and , to much lesser extent , to complex N-glycan chains . We next examined the binding of Tfp-coated green beads to glycan-coated red beads ( Figure 2E ) . Whereas Tfp-coated green beads mixed with BSA-coated red beads resulted in only 2% mixed yellow aggregates , incubation with N-glycan-3-coated red beads resulted in 74% yellow aggregates , which was decreased to 12% upon addition of excess N-glycan-3 ( Figure 2E ) . We also tested the ability of Tfp-coated green beads to interact with red beads coated with individual sugars . Mixing Tfp-coated green beads with Man-coated red beads resulted in only 25% yellow aggregates , and in 16% yellow aggregates when Tfp-coated green beads were mixed with Gal-coated red beads . Importantly , Tfp-coated green beads did not aggregate with HS-coated red beads . These findings confirm that Tfp preferentially interact directly with complex N-glycans in a polyvalent manner and individual sugars do not provide enough strength and specificity for the interaction . Taken together , in vitro binding assays conclusively show that Tfp are necessary and sufficient to interact with N-glycans , which corroborates our cell-culture based experiments . Our results thus far suggest that Tfp are necessary in vivo and sufficient in vitro to interact with N-glycans whereas flagella are necessary in vivo and sufficient in vitro to interact with HS . As it is possible that additional host molecules mediate bacterial binding to cultured cells , we tested by IF microscopy whether Tfp-coated or flagella-coated beads would be sufficient to mediate binding to the AP surface of airway epithelial cells and compared these results with the binding of PAO1ΔpilA and PAO1ΔfliC . For these experiments we utilized Calu-3 cells grown as confluent monolayers on Transwell filters for a shorter time period ( 3 days rather than the usual 9 days ) . Under these conditions , functional tight junctions are formed , but the polarized distribution of HSPGs is not complete , i . e . some HSPGs are still present at the AP surface [9] . These 3-day grown monolayers , which we term incompletely polarized , were briefly treated with heparinase III or tunicamycin to further decrease the surface presentation of HSPGs or N-glycans , respectively . The resulting patchy distribution of HSPGs or N-glycans at the AP surface allowed us to correlate and quantify by IF the binding of PAO1 flagella or Tfp mutants , or adhesin-coated beads , to HS-rich or -poor areas ( visualized with an antibody to HS ) or N-glycan-rich or -poor areas ( visualized with fluorescent lectin that binds to Man residues in N-glycan chains ) . We first examined the binding of GFP-expressing PAO1ΔpilA and PAO1ΔfliC to membrane regions rich in HS or N-glycan chains . As shown in Figures 3A and B , significantly more PAO1ΔpilA-GFP co-localized to HS-rich patches ( 65% ) than to HS-poor patches ( 35% ) on the AP surface of heparinase III-treated Calu-3 monolayers . In contrast , only ∼30% of PAO1ΔfliC-GFP bound to HS-rich patches whereas ∼70% co-localized with HS-poor patches . In Calu-3 cells briefly treated with tunicamycin , only ∼40% of PAO1ΔpilAI-GFP co-localized to N-glycan-rich patches whereas a much larger fraction , almost 70% , of PAO1ΔfliC-GFP co-localized to N-glycan-rich areas ( Figures 3C and D ) . In summary , PAO1ΔpilA , for which flagella serve as the major adhesin , preferentially co-localized with HS-rich areas . In contrast , PAO1ΔfliC , for which Tfp serve as the major adhesin , preferentially co-localized with N-glycan-rich areas at the AP surface of incompletely polarized epithelium . We then tested whether Tfp or flagella were sufficient to mediate these specific interactions by quantifying the binding of Tfp- or flagella- coated green fluorescent beads to N-glycan or HSPG-rich areas at the AP surface . As shown in Figures 3E and F , 75% of flagella-coated beads co-localized with HS-rich areas at the AP surface of heparinase III-treated Calu-3 cells . In stark contrast , only 15% of Tfp-coated beads co-localized with HS-rich patches . The opposite results were observed for binding to N-glycans: ∼30% of flagella-coated beads compared to 80% of Tfp-coated beads co-localized with N-glycan-rich areas ( Figures 3G and H ) . Importantly , the results paralleled what was observed with intact bacteria . Altogether , these experiments demonstrate that Tfp are sufficient to mediate P . aeruginosa binding to N-glycan chains and that flagella are sufficient to mediate P . aeruginosa binding to HS chains of HSPGs at the surface of airway epithelium . Following binding to the epithelium , P . aeruginosa is able to enter into non-phagocytic cells; this event is most readily detectable in ∼75% of clinical , environmental , and laboratory strains , including PAO1 or PAK , that do not secrete the potent phospholipase ExoU but that encode ExoS [31] , [41] . In order to examine the role of flagella and Tfp interactions during entry at the AP or BL surface , polarized Calu-3 cells were pre-treated with various agents and standard bacterial invasion assays were performed . In general , the results of the invasion assays with PAO1 , PAO1ΔpilA , and PAO1ΔfliC were directly proportional to adhesion assays ( see Figure 1 ) , i . e . , more binding resulted in more invasion . Competitive inhibition with heparin or enzymatic cleavage of HS by heparinase III reduced PAO1 and PAO1ΔpilA invasion at the BL , but not AP surface , of polarized cells ( Figures 4A and B ) . Up-regulation of N-glycosylation by long-term culture of Calu-3 cells in the presence of Man enhanced the internalization of PAO1 or PAO1ΔfliC at the AP surface of polarized cells , while inhibition of N-glycosylation with tunicamycin reduced bacterial entry at the AP surface of polarized Calu-3 cells ( Figures 4A and C ) . Notably , inhibition of N-glycosylation had a small but statistically significant effect on PAO1 and PAO1ΔpilA entry at the BL surface , although it did not have any effect on bacterial binding . This finding suggests that flagella-dependent entry , but not binding , at the BL surface may be mediated by a yet unidentified N-glycosylated receptor ( s ) . Together , these data confirm that P . aeruginosa-induced binding and subsequent entry into polarized epithelium are primarily mediated by Tfp-dependent binding to host N-glycans at the AP surface and by flagella-dependent binding to HS chains at the BL surface of polarized epithelium . We have previously shown that P . aeruginosa internalization at the AP surface of incompletely polarized MDCK cells is actin-dependent and requires activation of PI3K and its effector protein Akt [33] . We therefore determined whether N-glycans and/or HSPGs acted as host receptors upstream of this signaling pathway and whether bacterial Tfp and flagella were bacterial partners associated with the PI3K/Akt-dependent entry at the AP or BL surface of polarized lung airway epithelial cells . Pre-treatment with LY29004 , an inhibitor of PI3K , did not affect PAO1 binding at either the AP or BL surface of fully polarized Calu-3 cells ( Figure S4 ) . However , it had a pronounced effect on bacterial invasion at the BL surface , reducing it ∼5 fold , and it had a smaller but statistically significant effect on invasion at the AP surface ( Figure 4A . ) . PI3K-dependent entry at the BL surface required flagella binding to HS , as inhibition of PI3K , competitive inhibition with heparin , or heparinase-III treatment decreased PAO1ΔpilA entry similarly to PAO1 , but had no effect on the already low levels of PAO1ΔfliC entry ( Figures 4B and C ) . At the AP surface , inhibition of PI3K caused a small but statistically significant decrease in PAO1ΔfliC invasion . Simultaneous PI3K inhibition and tunicamycin treatment did not further reduce PAO1ΔfliC entry ( Figure 4C ) , suggesting that PI3K-dependent invasion at the AP surface could require Tfp-mediated binding to N-glycans . Together , these results suggest that flagella-mediated binding at the BL surface leads to P . aeruginosa internalization through a PI3K-dependent pathway . At the AP surface , Tfp mediated entry through a PI3K-dependent entry can also occur , although consistent with our previously published results [9] , BL entry is more efficient than AP entry . On the basis of our results , we would predict that flagella-mediated binding and entry at the BL surface or Tfp-mediated binding and entry at the AP surface should increase downstream Akt phosphorylation . To test this hypothesis , polarized Calu-3 cells were co-cultivated for 60 min with PAO1 or with adhesin mutants , and Akt was immunoprecipitated followed by immunoblotting with anti-phospho AktSer473 antibody . The ratio of phospho-Akt to total Akt was quantified and normalized to the ratio observed in untreated cells . In control experiments , Calu-3 cells were AP or BL exposed to heparin-binding EGF-like growth factor ( HB-EGF ) for 10 min since Akt phosphorylation is well established as a downstream consequence of EGFR activation . BL addition of HB-EGF increased Akt phosphorylation over 2-fold , but had little effect when added to the AP surface ( Figures 5A and B ) , which is consistent with the known BL localization of EGFR in polarized epithelium and , thus , BL activation of Akt . Addition of PAO1 or PAO1ΔpilA at the BL surface resulted in activation of Akt , with a 2 . 5-fold increase in the fraction of phosphorylated Akt when compared to bacterial addition to the AP surface ( Figures 5A–D ) . However , AP addition of PAO1ΔpilA or BL addition of PAO1ΔfliC failed to activate Akt ( Figures 5C–F ) . This observation is consistent with a requirement for initial Tfp-mediated binding at the AP surface and flagella-mediated binding at the BL surface . Inhibition of PI3K prior to the addition of bacteria almost completely eliminated Akt phosphorylation at the AP and BL surface ( Figures 5A–F ) , consistent with the known role of PI3K and Akt activation in P . aeruginosa entry [33] . We next determined whether Tfp or flagella-mediated activation of Akt involved HSPGs or N-glycans . Pre-treatment of Calu-3 cells with heparinase III reduced Akt phosphorylation to near basal levels upon BL , but not AP , addition of PAO1 or PAO1ΔpilA ( Figures 5A–D ) . Inhibition of N-glycosylation with tunicamycin partially reduced Akt activation upon AP addition of PAO1 and PAO1ΔfliC ( Figures 5A–B and E–F ) . Interestingly , inhibition of N-glycosylation reduced Akt phosphorylation upon BL infection with PAO1 or PAO1ΔpilA , which is consistent with our bacterial invasion results ( see Figure 4 ) and suggests involvement of yet unidentified N-glycosylated receptor in flagella-HS-dependent invasion and activation of PI3K/Akt pathway . Together , these results strongly suggest that activation of PI3K/Akt pathway at the BL surface is primarily dependent on flagella-mediated bacterial binding to HS chains of HSPGs , while at the AP surface it is dependent on Tfp-mediated bacterial binding to N-glycan chains . While many stimuli can activate the PI3K/Akt pathway , we were particularly interested in interrogating whether flagella- or Tfp-mediated binding to HSPGs or N-glycan chains and/or subsequent bacterial entry involved growth factor receptors ( GFRs ) . Notably , GFRs require N-glycosylation for their activity and many of them are also modulated by HSPGs; our work clearly establishes the role of both N-glycans and HSPGs for P . aeruginosa binding and internalization . In preliminary studies , we investigated the role of epidermal GFR ( EGFR ) , platelet-derived GFR ( PDGFR ) , and fibroblast GFR ( FGFR ) . While pharmacologic inhibition of any of the GFRs did not affect bacterial binding , inhibition of EGFR and PDGFR , but not FGFR , reduced bacterial internalization at the BL surface of Calu-3 cells ( Figure S4 ) . We also confirmed the role of EGFR and PDGFR in P . aeruginosa entry using siRNA gene silencing in HeLa cells ( Figure S4 ) . Simultaneous removal of HS by heparinase III and pharmacologic inhibition of EGFR had the same effect as heparinase III treatment alone ( Figure S4 ) . These results suggest that EGFR may potentially mediate bacterial entry upon HS-dependent bacterial binding , and thus EGFR was a logical candidate to study further . To elucidate the role of EGFR in flagella- and Tfp-mediated bacterial internalization , we first showed that inhibition of EGFR reduced invasion of PAO1 and PAO1ΔpilA at the BL surface of polarized Calu-3 cells , but did not further decrease the invasion of PAO1ΔfliC ( Figure 4 ) . Inhibition of EGFR did not have any effect on bacterial internalization at the AP surface since EGFR is predominantly expressed on the BL surface of polarized epithelium . Simultaneous inhibition of N-glycosylation and pharmacologic inhibition of EGFR reduced PAO1 and PAO1ΔpilA internalization at the BL surface similarly to what was observed with each treatment alone ( Figures 4A and B ) , confirming that EGFR activity depends on its N-glycosylation . Inhibition of PAO1ΔpilA entry by heparin , heparinase III treatment , or PI3K inhibition reduced bacterial entry to similar degrees , while EGFR inhibition had somewhat intermediate effect ( Figure 4B ) . Furthermore , combined heparinase III treatment and EGFR inhibition reduced PAO1 and PAO1ΔpilA internalization to a greater degree than inhibition of EGFR alone ( Figures 4A and B ) . These results suggest that bacterial internalization occurs through multiple HSPG-dependent pathways , including one that involves a flagella-HS-EGFR complex leading to PI3K activation at the BL surface of polarized epithelium . To further investigate the role of flagella or Tfp during EGFR-dependent entry at the AP or BL surface of polarized epithelial cells , we tested whether bacterial binding induced EGFR phosphorylation . Total EGFR was immunoprecipitated from cell lysates 1 h after AP or BL infection with PAO1 , PAO1ΔpilA or PAO1ΔfliC , followed by immunoblotting with a monoclonal anti-phospho EGFRSer1046/1047 antibody . The ratio of phospho-EGFR to total EGFR was quantified and normalized to the ratio observed in untreated cells . In control experiments , BL exposure of Calu-3 cells to HB-EGF for 10 min increased EGFR phosphorylation 3-fold , but had little effect when applied to the AP surface ( Figures 6A and B ) , consistent with the known BL localization of EGFR in polarized epithelium . Binding of PAO1 or PAO1ΔpilA to the BL surface of polarized epithelium resulted in a 2- to 2 . 5-fold increase in the fraction of phospho-EGFR ( Figures 6A–D ) . In contrast , binding of PAO1ΔfliC failed to increase EGFR phosphorylation above background levels ( Figures 6E and F ) , suggesting that flagella-mediated binding is required for EGFR activation . Activation of EGFR upon BL addition of PAO1 or PAO1ΔpilA was reduced by EGFR inhibition , removal of HS by heparinase III , or inhibition of N-glycosylation by tunicamycin , confirming the involvement of bacterial flagella , host HSPGs , and N-glycosylation in bacteria-mediated EGFR phosphorylation . Finally , we tested whether flagella-mediated PI3K activation was EGFR dependent . Indeed , inhibition of EGFR decreased Akt phosphorylation in response to BL addition of PAO1 or PAO1ΔpilA , but not PAO1ΔfliC ( Figures 5A–D ) . These results indicate that P . aeruginosa-mediated internalization at the BL surface occurs principally via flagella-mediated interactions through a pathway that utilizes HSPGs and that involves EGFR and PI3K/Akt activation . At the AP surface , Tfp mediates bacterial entry through a PI3K/Akt pathway that is independent of EGFR . To elucidate whether Tfp and/or flagella alone were sufficient to induce phosphorylation of EGFR , beads coated with purified adhesins were added to polarized Calu-3 cells for 1 h and immunoprecipitation/immunoblotting assays were performed as previously for the whole bacteria . Remarkably , addition of flagella-coated beads to Calu-3 cells increased the fraction of phospho-EGFR ( 1 . 7-fold ) ( Figures 7 C and D ) , almost to levels seen upon BL addition of PAO1 ( 2 . 2-fold ) ( Figures 6 A and B ) . In contrast , no increase in the ratio of phospho-EGFR was observed upon addition of the flagella-coated beads to the AP surface or upon AP or BL addition of BSA-coated beads . Similar to what we observed with PAO1 or PAO1ΔpilA , flagella-mediated phosphorylation of EGFR was reduced after inhibition of EGFR , HS removal by heparinase III , or inhibition of N-glycosylation by tunicamycin ( Figures 7C and D ) . Notably , AP or BL addition of Tfp-coated beads did not detectably increase phospho-EGFR above background levels ( Figures 7E and F ) . We were unable to detect induction of Akt phosphorylation by either flagella- or Tfp-coated beads ( Figures 7A and B ) . Altogether , these results demonstrate that flagella alone are sufficient to induce EGFR phosphorylation . To confirm the specificity and significance of EGFR phosphorylation triggered by flagella-coated beads , we tested other targets of EGFR activation . As shown in Figures 7G and H , addition of flagella-coated beads to the BL surface of polarized epithelium resulted in phosphorylation of EGFR adaptor protein Shc ( Src Homology-2 Domain Containing Transforming Protein ) [42] . Shc exists in three isoforms ( p46 , p52 , and p66 ) and we detected elevated levels of phosphorylated p46 and p52 isoforms upon infection with PAO1 or flagella-coated beads . In contrast , there was no increase in the ratio of phospho-Shc upon addition of the flagella-coated beads to the AP surface or upon AP or BL addition of pili-coated beads . Although , at this point , we are not able to show the activation of signaling events farther downstream of EGFR , such as Akt phosphorylation , these results confirm the significance of EGFR phosphorylation by flagella- but not pili-coated beads . HB-EGF , but not EGF , needs to bind to HS chains of HSPGs to activate EGFR . Since our results demonstrate that flagella likewise activate EGFR in an HS-dependent manner , we first tested whether flagella-coated beads bind to EGFR or HB-EGF . While flagella-coated beads bound with great avidity to HS coated onto plastic wells , they did not bind measurably to the extracellular domain of EGFR , to HB-EGF , or to EGF , used as a negative control since it does not bind to HS to activate EGFR ( data not shown ) . Second , we tested whether HB-EGF can inhibit binding of flagella-coated beads to HS chains . At a high concentration , exogenous HB-EGF , but not EGF , slightly inhibited binding of flagella-coated beads to HS coated onto plastic wells ( Figure S5 ) . These results suggest that , at high concentrations , HB-EGF can either compete with bacterial flagella for binding sites on HS chains or it sterically hinders flagella binding to HS .
Adhesion of pathogens to the host epithelium is an early and critical step in mucosal infections , and successful pathogens exploit specific niches to colonize , obtain nutrients , replicate , and disseminate . In previous studies utilizing well polarized epithelial cells , we have shown that the important nosocomial pathogen P . aeruginosa binds preferentially to different host molecules at the AP versus BL surface , specifically to N-glycans at the AP surface and to HSPGs at the BL surface [9] . We hypothesized that these complex P . aeruginosa-host interactions may be mediated by distinct bacterial adhesins . In the current studies , we identify the bacterial adhesins that are necessary and sufficient to mediate these different binding specificities . We demonstrate that Tfp are necessary to mediate maximal binding and entry at the AP surface through N-glycans , while flagella are required to mediate maximal binding and entry through HSPGs at the BL surface of polarized epithelium . While flagella have been shown previously to be required for the host response to BL infection with P . aeruginosa [43] , [44] , our studies using beads coated with purified Tfp or flagella are the first to demonstrate that Tfp or flagella are sufficient to mediate the differential binding to N-glycans or HSPGs , respectively . Our studies reveal that flagella can also mediate , to a small extent , bacterial binding to N-glycans at the AP surface of polarized epithelium . However , we cannot determine at this point how significant these interactions are in P . aeruginosa infections due to the constrains of utilized assays; advanced more sensitive assays need to be employed to further investigate these interactions . Although we can detect N-glycan-dependent bacterial binding at the BL surface , neither up- nor down-regulation of N-glycosylation has any effect on the binding . Therefore , we postulate that binding of P . aeruginosa to N-glycan chains on the BL surface of polarized epithelium is not essential , rather the protein core of N-glycoprotein ( s ) play a role in the binding . We investigated the consequences of these binding events on induction of host cell signal transduction pathways . We find that flagella-dependent bacterial binding to HS at the BL surface of polarized epithelium leads to activation of EGFR , as evidenced by increased phosphorylation of EGFR and of two of its associated downstream targets , the adaptor protein Shc and the serine/threonine kinase Akt . Strikingly , flagella alone are sufficient for EGFR and Shc phosphorylation upon binding of flagella-coated beads to the BL surface . Tfp-mediated bacterial binding at the AP surface also results in increased Akt phosphorylation , although the activation is less robust when compared to flagella-mediated bacterial binding at the BL surface of polarized epithelium . Although we were able to detect elevated levels of phosphorylated EGFR and Shc upon binding of flagella-coated beads to the BL surface , we were unable to detect induction of Akt phosphorylation by either flagella- or Tfp-coated beads . Adhesin-coated beads may not be sufficient to trigger more downstream signaling events and additional bacterial factors may be required [31] , [41] . Furthermore , adhesins coated onto beads may not be fully functional . Although we show that three-dimensional structure of isolated pili is most likely intact , Tfp extension/retraction is compromised and , thus , certain Tfp functions may be hindered when studied in the context of adhesin-coated beads . Finally , we cannot exclude that our detections assays are not sensitive enough to measure induction of downstream signaling events . Based on our results , we propose a model ( Figure 8 ) , in which P . aeruginosa binds in a Tfp-dependent manner to N-glycan chains of one or more yet unidentified glycoproteins , which leads , to some extent , to the activation of PI3K/Akt pathway at the AP surface of polarized epithelium . It remains to be investigated whether other signaling pathways are also activated upon Tfp-mediated binding to N-glycans . At the BL surface , P . aeruginosa first binds in a flagella-dependent manner to HS chains of HSPGs . An attractive model is that a complex is formed between flagella , HPSGs , HB-EGF , and EGFR , which then leads to activation of EGFR and subsequent activation of the PI3K/Akt pathway . Flagella-mediated activation of EGFR most likely involves initial binding of flagella to HS chains of HSPGs since flagella-coated beads bind with great avidity to HS , but they do not bind measurably to HB-EGF or to the extracellular domain of EGFR . We attempted to discern whether flagella and HB-EGF compete for binding to HS; however , only a high concentration of an exogenous HB-EGF , far greater than concentrations required for EGFR phosphorylation in vitro , interferes with flagella binding to HS . Finally , we cannot exclude that other signal transduction pathways are activated upon flagella mediated binding to HSPGs , independent and/or dependent on EGFR phosphorylation . One consequence of activation of EGFR and the PI3K/Akt pathway is P . aeruginosa internalization into host epithelium . Interestingly , inhibition of PI3K has a more pronounced effect on bacterial internalization at the AP surface of incompletely polarized epithelial cells ( unpublished data and [33] ) when compared to well polarized cells here studied . This phenomenon likely reflects differences in the composition of the AP versus BL surface during different stages of cell polarization . There may be increased levels of HSPGs on the AP surface of incompletely polarized cells [9]; in addition , there may be differences in how P . aeruginosa-mediated pathogenic events are affected by changes in the levels of host receptors that occur during the polarization process ( unpublished data and [9] ) . Bacterial binding and activation of EGFR and the PI3K/Akt pathway most likely lead to other pathogenic events as well . We have recently shown that inhibition of bacterial binding to N-glycans at the AP surface and to HS at the BL surface reduces P . aeruginosa-mediated host damage [9] . Similarly , pharmacologic inhibition of EGFR reduces bacterial cytotoxicity at the BL surface and inhibition of PI3K reduces bacterial toxicity at both the AP and BL surface of polarized epithelium ( unpublished data ) . However , it has been shown that P . aeruginosa infection of incompletely polarized corneal epithelial cells that leads to EGFR activation through shedding of the HB-EGF ectodomain , followed by activation of ERK1/2 and PI3K pathways , results in inhibition of apoptosis in the early stage of bacterial infection [45] . Thus , further studies are needed to elucidate the role of HSPGs- and HB-EGF-dependent EGFR and PI3K/Akt pathways in P . aeruginosa-mediated cell death . Activation of EGFR and other growth factors is an emerging theme in bacterial pathogenesis . Multiple pathogens have been shown to activate EGFR , including Neisseria gonorrhea , Neisseria meningitides , Helicobacter pylori , Pasteurella multocida , and Haemophilus influenzae [46]–[52] . Of particular relevance are studies with N . gonorrhea , which , similar to P . aeruginosa , binds to the AP surface of polarized epithelial cells as microcolonies that initiate changes in the host cell actin cytoskeleton and allow the microcolonies to enter into epithelial cells . Upon binding , EGFR is phosphorylated , its activation is required for N . gonorrhea internalization , and phospho-EGFR is found in close apposition to a fraction of surface bound microcolonies [49] , [52] . We were unable to determine by IF microscopy whether phospho-EGFR co-localized with bound flagella-expressing P . aeruginosa or with flagella-coated beads , because of high background from staining with the anti-phospho-EGFR antibody . Nonetheless , the similarities between these two organisms are intriguing . Our finding of Tfp- and flagella-dependent binding to N-glycans at the AP surface and HSPGs at the BL surface of polarized epithelium , respectively , and subsequent EGFR/PI3K/Akt signaling events introduce another level of complexity to diverse mechanisms of P . aeruginosa adhesion and establishment of an infection . Several N-glycosylated receptors have been identified , including the Cystic Fibrosis Transmembrane Regulator ( CFTR ) , fibronectin , or integrins [53] , [54] . However , since fibronectin and integrins are preferentially expressed at the BL surface , their N-glycan chains are unlikely to be AP receptors for bacterial Tfp . Although CFTR is expressed on the AP surface of polarized epithelium [53] , it is also unlikely that N-glycan chains of CFTR mediate Tfp-dependent bacterial binding under the conditions of our experiments , as we observed similar levels of bacterial adhesion to the AP surface of epithelial cells that express either very low or high amounts of CFTR ( [9] and unpublished data ) . Previous studies have also suggested that glycosphingolipids , i . e . , asialoGM1 , may serve as AP receptors for Tfp [22] , [23]; however , glycosylation of sphingolipids differs from N-glycosylation and , thus , Tfp binding to N-glycans characterized in this paper most likely represents a distinct mechanism by which P . aeruginosa is able to infect the host . Flagella , likely through its interaction with TLR5 , have been shown to activate the innate immune response preferentially at the basolateral surface of polarized airway epithelial cells [43] , [44] . Utilizing informative flagella mutants or purified flagellin , it has been possible to uncouple TLR5-mediated NFκB-dependent inflammatory responses from EGFR-dependent epithelial cell proliferation , wound repair , and antimicrobial peptide production [55] . TLR5 activates EGFR through signaling events that are not dependent on HSPGs [56] and , thus , TLR5 is very unlikely to be involved in flagella- and HSPGs-dependent cascade leading to EGFR phosphorylation . TLR5 is predicted to be N-glycosylated [57] and since we show that modulation of N-glycosylation does not affect flagella-mediated binding to the BL surface , N-glycan chains of TLR5 most likely do not mediate binding of bacterial flagellin . Since expression of most TLRs can be variable [58] and we can detect low flagella-dependent P . aeruginosa binding to N-glycans on the AP surface of polarized epithelium , we cannot exclude that in certain pathogenic conditions P . aeruginosa may bind in a flagella-dependent manner to N-glycan chains of TLR5 on the AP surface . Other bacterial factors have been implicated in mediating P . aeruginosa binding to the host . P . aeruginosa LPS binds to TLR4 , predominantly expressed on the BL surface of polarized epithelium , and it has been reported to stimulate human lung epithelial wound repair through a TLR4- and EGFR-dependent pathway that involves release of the EGFR ligand , TGF-α , by the matrix metalloprotease TACE [59] . Flagellar components have been shown to bind to LewisX derivatives found on secreted mucins [19] and P . aeruginosa can additionally stimulate mucin secretion in an EGFR-dependent manner , as shown in rat tracheal cells [60] . Furthermore , two different P . aeruginosa lectins , PA-IL ( LecA ) and PA-IIL ( LecB ) , have been implicated in bacterial binding to sugar moieties present on mucins or cell surface receptors [36] . Work from our lab and others have implicated more receptors and signaling pathways in P . aeruginosa entry or host responses to bacterial infection , including PDGFR , Abl/Crk [31] , and Src family kinases [61] , e . g . Lyn [62] , and it will be of interest to determine if any of these molecules are differently activated upon Tfp- or flagella-mediated binding at the AP or BL surface of polarized epithelium . In summary , P . aeruginosa can utilize numerous adhesins or virulence factors and exploit numerous host receptors to adhere to the host epithelium . Studies focused on providing key insights into multifactorial and complex P . aeruginosa binding are crucial for comprehensive understanding of this event and identification of potential therapeutic targets . Our findings introduce bacterial and host players and link them with previously described signaling events to build novel network of interactions and events that lead to establishment of P . aeruginosa acute or chronic infections . Tfp and flagella as well as corresponding glycosylated host receptors are potentially valuable targets for designing therapies that interfere with the initial steps in P . aeruginosa infection and colonization . Such therapies could also target a number of other carbohydrate-based interactions of P . aeruginosa with the host , including bacterial binding to mucus and biofilm formation [32] , [36] . Therefore , these anti-adhesive therapies are a very attractive strategy for development of novel treatments for a wide range of both acute and chronic P . aeruginosa infections .
P . aeruginosa strain O1 ( PAO1 ) was obtained from the ATCC ( ATCC 15692 ) and isogenic mutants PAO1ΔpilA and PAO1ΔfliC were previously constructed in the laboratory [63] . PA103 [64] was a kind gift of Dr . Dara Frank and PA103 Mutant 9 was previously constructed in the laboratory [38] . All strains were routinely grown shaking overnight in Luria-Bertani broth ( LB broth ) at 37°C . GFP-expressing strains were created by electroporation of the pnpT2-GFP-pUCP20 plasmid as described previously [9] . Following overnight growth in LB broth at 37° with shaking , bacteria were added to well polarized cells at an MOI of 20 . For AP infections , the bacteria in serum-free MEM were added to the AP chamber of cells grown on Transwells . For BL infections , the Transwell insert was placed directly onto 50 µl of serum-free MEM containing bacteria . After 2 h of infection at 37°C , adhesion and invasion assays were performed as described previously [9] . Bacteria were enumerated by plating serial dilutions of cell lysates to LB plates and counting colony-forming units ( cfu ) . All assays were carried out on triplicate wells , and results are reported as the average of three to five experiments . To isolate surface flagella or Tfp , PAO1ΔpilA , PAO1ΔfliC , PA103 , or PA103 Mutant 9 were grown shaking overnight in 2 ml LB at 37°C and 100 µl of the culture was plated onto 1 . 5% LB agar . Following growth at 37°C for 16 h , the bacteria were scraped from the agar surface and resuspended in 5 ml PBS . A volume of cells equivalent to an optical density at 600 nm ( OD600 ) of 20 . 0 was resuspended in 1 ml PBS . Cells were vortexed vigorously at room temperature for 30 min to remove surface flagella or Tfp by shearing . The suspension was centrifuged at 20 , 000× g for 10 min at 4°C , the supernatant was collected and centrifuged a second time to remove all cellular debris . The resulting supernatant was dialyzed against PBS ( pH 7 . 4 ) overnight at 4°C and centrifuged at 20 , 000× g for 20 min at 4°C to remove insoluble proteins . Afterward , the supernatant was incubated overnight at 4°C in 100 mM MgCl2 to precipitate flagella or Tfp . The precipitate was collected by centrifugation at 4°C ( 15 , 000× g for 20 min ) , and the pellet was resuspended and dialyzed against PBS ( pH 7 . 4 ) overnight at 4°C . Afterward , the suspension was centrifuged at 20 , 000× g for 20 min at 4°C to remove insoluble proteins , and flagella or Tfp were precipitated again in 100 mM MgCl2 at 4°C . Dialysis , centrifugation , and precipitation steps were repeated again to obtain flagella or Tfp of a high purity as assessed by SDS-PAGE and staining in 0 . 25% Coomassie ( Bio-Rad ) for 4 h . Destaining was done in 10% ethanol , 7 . 5% acetic acid for 6 h or until bands appeared and the background was clear . For adsorption of bacterial adhesins to beads , 0 . 5 ml of 2 . 5% suspension of Green Fluoresbrite Latex fluorescent beads ( 2 µm ∅; Polysciences Inc . ) were mixed with 200 µg of purified Tfp or flagella in 0 . 1 M Borate Buffer overnight at room temperature according to the manufacturer's protocol . To determine the coating efficiency , coated beads were eluted in SDS sample buffer and analyzed by Western blotting . Following SDS-PAGE and transfer , the membranes were probed with a 1∶100 , 000 dilution of primary α-FliC ( for flagella ) or α-PilA ( for Tfp ) [63] antibody overnight at 4°C , followed by probing with a 1∶25 , 000 dilution of horseradish peroxidase-conjugated secondary antibody ( Jackson ImmunoResearch Laboratories ) . Gels were quantified by densitometry using ChemiDoc XRS and coating efficiencies were calculated . On average , 15–20% ( 30 µg ) flagella and ∼30% ( 60 µg ) Tfp were bound to the beads . For adsorption of glycosylated molecules to beads , 0 . 5 ml of Red Fluoresbrite Latex fluorescent beads ( 2 µm ∅; Polysciences , Inc . ) were mixed with 400 µg of HS , N-glycan-3 , Man , or Gal according to the manufacturer's protocol . Bead coating efficiency ( 30–40% ) was determined by eluting coated beads in SDS sample buffer , dotting an aliquot on a Zeta-Probe membrane ( Bio-Rad ) , staining the membrane with 1% Toluidine Blue , and comparing the staining to standards . Calu-3 cells were obtained from the ATCC and maintained in MEM supplemented with 10% fetal bovine serum ( FBS; Invitrogen ) and L-glutamate at 37°C with 5% CO2 . Cells were grown as 2D monolayers on 12-mm Transwell filters ( 3-µm pore size; Corning Inc . ) as previously described [9] . For experiments , Calu-3 cells were seeded at 1 . 5×106 cells/well and cultured for 3 days ( “incompletely polarized monolayers” ) or at 1×106 cells/well on Transwells and cultured for 9 days ( “well polarized monolayers” ) . Under each condition , cells formed polarized confluent monolayers as determined by polarized distribution of some AP and BL membrane proteins and the presence of functional tight junctions that were impermeable to small molecules such as FITC-inulin ( data not shown and [9] ) . However , in incompletely polarized monolayers , distribution of HSPGs on the BL surface was not fully polarized [9] . To remove glycosaminoglycans , Calu-3 cells were treated with 200 mU of heparinase III or chondroitinase ABC ( Sigma-Aldrich ) in Hank's Buffered Salt Solution ( HBSS ) containing 0 . 1% BSA at 37°C for 2 h ( resulting in 60–65% reduction in glycosaminoglycan expression ) , or , for partial removal of glycosaminoglycans , with 50 mU of heparinase III for 30 min ( resulting in ∼25% reduction in expression ) . To assess the efficiency of treatments , HS chains were visualized by IF staining with HS antibody ( 10E4; Seikagaku ) , CS chains were visualized with FITC-WFA ( CS-specific lectin from Wisteria floribunda; Sigma-Aldrich ) , and staining densities were quantified using ImageJ and compared to the staining densities of untreated cells ( data not shown and [9] ) . For competition blocking experiments with glycosaminoglycans , cells were pre-incubated with 5 µg/ml of heparin or CS ( Sigma-Aldrich ) at 37°C for 1 h in serum-free MEM . For up-regulation of N-glycosylation , cells were grown in the presence of 1 mM Man or Gal ( Sigma-Aldrich ) in MEM with 10% FBS for 1 week ( resulting in 1 . 4-to-1 . 7 fold increase in N-glycosylation ) . To inhibit N-glycosylation , cells were pre-treated with 1 µg/ml of tunicamycin ( Sigma-Aldrich ) for 16 h ( ∼50% reduction ) or for 8 h ( brief de-glycosylation resulting in ∼20% reduction ) in MEM with 10% FBS . To assess cell surface N-glycosylation , cells were stained with the Man-specific lectin FITC-concanavalin A ( Sigma-Aldrich ) , staining densities were quantified using ImageJ and compared to the staining densities of untreated cells ( data not shown and [9] ) . To inhibit EGFR , PDGFR , or FGFR , cells were pre-incubated with 10 µM AG1478 , AG1296 , or PD173074 ( Calbiochem ) in MEM with 10% FBS for 1 h . To inhibit PI3K , cells were pre-incubated with 50 µM LY294002 ( Sigma-Aldrich ) in MEM with 10% FBS for 1 h . Inhibition efficiencies were quantified by Western blotting using phospho-specific antibodies . EGFR ( sc-29301 ) , PDGFR-α ( sc-29443 ) , PDGFR-β ( sc-29442 ) , and control ( sc-37007 ) siRNAs were purchased from Santa Cruz Biotechnology . HeLa cells ( ATCC CCL-2 ) , grown in MEM supplemented with 10% FBS , were transfected with siRNAs according to the manufacturer's instructions . After 42 h , standard adhesion and invasion assays were performed . In parallel , lysates were immunoblotted with appropriate antibodies to evaluate the efficiency of protein depletion . HSPGs , glycosaminoglycans , N-glycan-1 , -2 , and -3 , and sugars ( Sigma Aldrich; 0 . 1–10 µg in 0 . 2 ml ddH2O ) were added to 96-well polystyrene plate ( Corning ) and incubated overnight at 37°C until evaporated . Wells were washed with ddH2O and blocked in 0 . 1% BSA for 0 . 5 h at room temperature . Bound molecules were stained with 1% Toluidine blue ( Sigma-Aldrich ) and absorbance was measured at 630 nm . The absorbance of known concentration of molecules was used as the standard curve and the concentration of bound molecules ( µg/well ) was calculated . 100 µl of flagella- or Tfp-coated beads in ddH2O were added to coated wells and incubated for 2 h on a rotary shaker . Non-adherent beads were removed by washing with ddH2O . For some experiments , 5–20 ng/ml HB-EGF or EGFR were added to heparin-coated wells before addition of flagella-coated beads . Bound beads was quantified using a SpectraMax 340PC plate reader using SOFTmaxPro software ( Molecular Devices ) at λex = 480 nm and λem = 530 nm . Beads bound to non-coated wells were used as a control and subtracted out as background . The results are reported as the average of six experiments , each with at least 6 replicates . For the bead-bead aggregation assay , red fluorescent beads , 100 µl of green fluorescent beads coated with flagella or Tfp were allowed to aggregate with 100 µl of red fluorescent beads coated with various glycosylated molecules on a rotary shaker at 50 rpm for 2 h in ddH2O . For competition blocking experiments , 5 µg/ml of heparin or N-glycan-3 ( Sigma-Aldrich ) were added to wells . Images of aggregates were acquired with a confocal microscope ( LSM 510; Carl Zeiss MicroImaging , Inc . ) equipped with a 20× objective . Image processing was performed using Adobe Photoshop CS4 version 11 . 0 . 2 . Quantification of the fraction of green , red , and mixed ( yellow ) aggregates from 3 separate experiments and 10 events per each sample was performed using UTHSCSA Image Tool version 2 . 00 Alpha . Well polarized Calu-3 cells grown on Transwells for 9 days were washed and placed in serum-free MEM for ∼17 h . Bacteria at the MOI of 200 or 50 µl adhesin-coated beads were added to the AP or BL chamber for 1 h . As a control , 10 ng/ml HB-EGF was added to the AP or BL-chamber for 10 min . The infected and HB-EGF-treated monolayers were washed with cold PBS containing 1 mM sodium orthovanadate ( Sigma-Aldrich ) . Cells were lysed in modified radioimmunoprecipitation ( RIPA ) buffer ( 50 mM Tris , pH 7 . 4 , 150 mM NaCl , 2 mM EDTA , 2 mM EGTA , 1% Triton X-100 , 0 . 5% deoxycholate , 0 . 1% SDS , 1 mM sodium orthovanadate , 50 mM NaF , 0 . 1 mM okadaic acid ( Sigma-Aldrich ) , 1 mM phenylmethylsulfonyl fluoride ( Sigma-Aldrich ) , and proteinase inhibitor tablets ( Complete; Roche Diagnostics ) ) for 20 min , and cells were removed from the Transwell filters by gentle scraping . The cell lysates were centrifuged at 16 , 000× g for 20 min . Immunoprecipitation with Akt or EGFR antibody ( Cell Signaling Technology ) using Magnetic Dynabeads Protein G beads ( Invitrogen ) were performed according to the manufacturer's protocol . For detection of Shc , whole cell lysates were used without immuprecipitation . Cell lysates or eluted immune complexes were separated on Novex-NuPAGE 10% Bis-Tris SDS-PAGE gels ( Invitrogen ) and electroblotted to iBlot Nitrocellulose Membranes using the iBlot Device ( Invitrogen ) according to the manufacturer's protocol . Membranes were blocked in PBS containing 0 . 05% Tween 20 and 5% non-fat milk ( PBST ) and probed with a 1∶1000 dilution of an antibody that recognizes Akt phosphorylated on serine 473 , EGFR phosphorylated on serine 1046/1047 , or Shc phosphorylated on tyrosine 239/240 ( Cell Signaling Technology ) in PBST , overnight at 4°C . Membranes were then incubated with a 1∶3000 dilution of horseradish peroxidase-conjugated secondary antibody ( Jackson ImmunoResearch Laboratories ) and detected by enhanced chemiluminescence using the Amersham ECL Western blotting detection kit ( GE Healthcare ) . For loading control , membranes were stripped and re-probed with an antibody that recognizes all forms of Akt , EGFR , or Shc ( Cell Signaling Technology ) . Gels were quantified by densitometry using ChemiDoc XRS . Calu-3 cells grown on Transwells as incompletely polarized monolayers were infected with PAO1ΔpilA-pGFP , PAO1ΔfliC-GFP ( MOI 50 ) , or with 50 µl flagella- or Tfp-coated green fluorescent beads for 2 h at room temperature . Afterwards , cells were washed and fixed in PBS containing 1% paraformaldehyde at 37°C for 0 . 5 h . After washing , cells were incubated with primary antibodies overnight at 4øC and , afterwards , with fluorescent secondary antibodies for 2 h at room temperature . HS chains were stained with 1∶500 anti-heparan sulfate antibody ( 10E4; Seikagaku ) followed by 1∶2 , 000 AlexaFluor647-conjugated secondary antibody ( Invitrogen ) . Actin filaments were stained with 1∶2 , 000 AlexaFluor594-phalloidin ( Invitrogen ) and Man residues were detected by staining with 1∶1 , 000 FITC-conjugated lectin concanavalin A ( Sigma-Aldrich ) . Filters were excised and mounted on microscope slides ( Fisher Scientific ) in mounting medium ( Vector Laboratories , Inc . ) . Samples from 3 separate experiments and 20 events per each sample were examined with a confocal microscope ( LSM 510; Carl Zeiss MicroImaging , Inc . ) . Images and 3D reconstructions were acquired by and processed in Meta 510 software . Image J analysis was performed on TIFF files . Bacterial or bead binding to Calu-3 cells and co-localization with surface markers was quantified using the Image J plugin Voxel counter on 3D reconstructions of TIFF images acquired with Meta 510 software . Voxel Counter ( ImageJ plugin ) was used to quantify the volume of bound 3D bacterial or bead aggregates and a minimum volume was set as a threshold to enable automated cell counting using the 3D Object Counter ( ImageJ plugin ) . Any aggregate above the threshold was counted as one . The surface area of membrane regions either enriched or depleted of HS or N-glycans ( as determined by staining with an anti-HS antibody or with FITC-ConA , respectively ) was measured in pixels by ImageJ , and the number of bacterial or bead aggregates bound was normalized per pixel of each specific surface area . The percentage ( compared to total ) of bacterial or bead aggregates bound to each specific region was determined . Data are expressed as means ± SD ( standard deviation ) . Statistical significance was estimated by ANOVA test using InStat version 3 . 0b . Differences were considered to be significant at P<0 . 05 . | Pseudomonas aeruginosa is one of the most virulent nosocomial opportunistic pathogens that is associated with a broad spectrum of life-threatening infections . Antibiotic resistance is widespread and attributable mortality remains near 50% . Complex binding to epithelial cells is a key first step for this potent pathogen to unleash its armamentarium of virulence factors . Polarized epithelium has distinct apical ( AP ) and basolateral ( BL ) surface , composed of different glycosylated molecules , and P . aeruginosa can potentially employ different adhesins to bind to these receptors . Using isogenic mutants as well as in vitro cell-free assays , we demonstrate that bacterial type IV pili are necessary and sufficient to mediate AP interactions with N-glycans whereas bacterial flagella interact with heparan sulfate chains of proteoglycans on the BL surface . These interactions induce specific host signaling pathways that lead to subsequent pathogenic events , such as bacterial entry into host epithelium . Moreover , we show that flagella alone are sufficient to activate the epidermal growth factor receptor and the adaptor protein on the BL surface . These studies reveal new information about key players in the versatile interactions of P . aeruginosa with the host and provide appealing targets for blocking early binding steps essential for establishment of P . aeruginosa infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biochemistry",
"infectious",
"diseases",
"genetics",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Pseudomonas aeruginosa Pili and Flagella Mediate Distinct Binding and Signaling Events at the Apical and Basolateral Surface of Airway Epithelium |
Pore-forming toxins ( PFTs ) are by far the most abundant bacterial protein toxins and are important for the virulence of many important pathogens . As such , cellular responses to PFTs critically modulate host-pathogen interactions . Although many cellular responses to PFTs have been recorded , little is understood about their relevance to pathological or defensive outcomes . To shed light on this important question , we have turned to the only genetic system for studying PFT-host interactions—Caenorhabditis elegans intoxication by Crystal ( Cry ) protein PFTs . We mutagenized and screened for C . elegans mutants resistant to a Cry PFT and recovered one mutant . Complementation , sequencing , transgenic rescue , and RNA interference data demonstrate that this mutant eliminates a gene normally involved in repression of the hypoxia ( low oxygen response ) pathway . We find that up-regulation of the C . elegans hypoxia pathway via the inactivation of three different genes that normally repress the pathway results in animals resistant to Cry PFTs . Conversely , mutation in the central activator of the hypoxia response , HIF-1 , suppresses this resistance and can result in animals defective in PFT defenses . These results extend to a PFT that attacks mammals since up-regulation of the hypoxia pathway confers resistance to Vibrio cholerae cytolysin ( VCC ) , whereas down-regulation confers hypersusceptibility . The hypoxia PFT defense pathway acts cell autonomously to protect the cells directly under attack and is different from other hypoxia pathway stress responses . Two of the downstream effectors of this pathway include the nuclear receptor nhr-57 and the unfolded protein response . In addition , the hypoxia pathway itself is induced by PFT , and low oxygen is protective against PFT intoxication . These results demonstrate that hypoxia and induction of the hypoxia response protect cells against PFTs , and that the cellular environment can be modulated via the hypoxia pathway to protect against the most prevalent class of weapons used by pathogenic bacteria .
Pore-forming toxins ( PFTs ) are by far the most abundant and amongst the most important bacterial protein virulence factors [1] . These toxins , secreted by many pathogenic bacteria , function by binding to receptors on host cell plasma membrane , oligomerizing , inserting , and then forming holes [2] . The importance of PFTs in promoting bacterial pathogenesis has been demonstrated by numerous experiments where individual PFTs have been genetically deleted from pathogenic bacteria and the bacteria then tested for reduced virulence [3] . Examples of PFTs that contribute significantly to bacterial virulence include α-toxin by Clostridium septicum , streptolysins by Group A and B Streptococci , α-toxin by Staphylococcus aureus , Vibrio cholerae cytolysin ( VCC ) , α-hemolysin from uropathogenic E . coli , cytolysin from Enterococcus faecalis , and crystal ( Cry ) proteins from Bacillus thuringiensis ( Bt ) . Although they are expressed by many bacterial pathogens and are broadly important as potentiators of infection , the effects of these toxins on host cells have been vastly understudied . There are several reasons for this lack of attention . First , their mechanism of action is deceptively simple . Second , most of the attention has been given to understanding how PFTs can change conformation from secreted , soluble proteins to insoluble proteins embedded in the plasma membrane . Third , because breaching of the plasma membrane is a major insult to a cell , a multitude of cellular responses to PFTs have been reported , including Ca2+ influx , K+ efflux , increased endocytosis/exocytosis , vacuolization , necrosis , and apoptosis [3] , [4] , [5] , [6] . Because the responses are so large and extensive , it has been daunting to determine whether these responses contribute to defense , intoxication , both , or neither . Fourth , most of the studies carried out to date involved cells in isolated culture , which does not always accurately recreate the response of cells to toxins in the context of intact tissue . To address many of these limitations , an excellent genetic system for studying PFT responses in an intact animal has recently emerged: the Bacillus thuringiensis ( Bt ) Crystal ( Cry ) PFT – Caenorhabditis elegans toxin-host interaction system [7] . C . elegans has become an important genetically tractable organism for studying immune responses to bacterial pathogens [8] . Bt is thought to be a natural pathogen of C . elegans [9] , [10] , [11] and is famous for the production of three-domain PFTs that are widely used in insect biocontrol [12] . The interaction of Cry proteins with C . elegans allowed for the first molecular PFT defense pathway identified , p38 mitogen-activated protein kinase ( MAPK ) pathway [13] . Loss of the p38 MAPK pathway was shown to result in loss of protection against Cry PFTs in C . elegans and was subsequently shown to result in loss of protection against PFTs in mammalian cells as well [13] , [14] . This same system was used to discover that the unfolded protein response ( UPR ) is also required for PFT defenses as a downstream target of the p38 MAPK pathway [15] . Both the p38 MAPK pathway and the UPR were demonstrated to be activated by PFTs in C . elegans and mammals [15] , [16] . Apart from these studies , only one other study to date has demonstrated a specific molecular pathway as involved in PFT responses [17] . Since , when studying intracellular PFT response pathways in the past , we have screened for C . elegans mutants hypersensitive to PFTs [13] , [15] , we reasoned that we could learn something different by screening for the opposite phenotype– C . elegans mutants resistant to PFTs . The reason for this assumption is that no intracellular pathway mutants were known that can make cells resistant to PFTs in general . Here we report on the results of a PFT resistance screen and find , unexpectedly , that resistance can be achieved by mutations that up-regulate the C . elegans low oxygen ( hypoxia ) response . Elimination of HIF-1 ( hypoxia inducible factor 1 ) , the main effector of the hypoxia pathway , abrogates this resistance and can lead to PFT hypersensitivity . This protection applies to multiple different PFTs and is clearly distinguished from the role of the HIF-1 pathway in other stress responses and aging . Furthermore , the hypoxia pathway is activated in response to PFTs , and low oxygen is itself protective against PFT attack . Our results indicate that the hypoxia/low oxygen response is likely to be of general importance for cellular responses to small-pore PFTs .
To identify pathways important for cellular responses to PFTs , we screened for mutants resistant to the PFT Cry protein , Cry21A . Cry21A is a three-domain Cry protein that targets nematodes and is in the same family as Cry5B [11] . Like Cry5B [18] , secondary structure programs predict Cry21A contains all the α helical segments that are involved in pore-formation in three-domain Cry proteins [19] . All three-domain Cry proteins , like Cry5B and Cry21A , are believed to act as PFTs , and pore-forming activity has been demonstrated for all Cry proteins so studied to date [12] . In the past , we have screened for mutants resistant to Cry5B , which has given rise to detailed understanding of the Cry5B receptor [20] , [21] , [22] but not to information on intracellular responses to Cry PFTs . Our rationale for screening for Cry21A PFT resistant animals was that it could elucidate new information about how cells respond to PFTs since Cry5B resistant mutants are only weakly resistant to Cry21A . To perform this screen , C . elegans hermaphrodites were mutagenized with EMS and allowed to self-fertilize for two generations . Sixty eight thousand F2 mutagenized hermaphrodites were fed an intoxicating dose Cry21A PFT and then screened for those that resisted intoxication . One mutant line , allele ye49 , was identified that bred true and is resistant to Cry21A PFT produced either from Bt or E . coli ( Figures 1A and 1B ) . To identify the gene mutated in ye49 , we performed standard three-factor and single-nucleotide mapping experiments , which narrowed the search to a region on chromosome V , between markers snp_F15H10 and snp_T21C9 , that includes 43 genes . Mutation in one of the genes in this region , egl-9 , had been previously identified as resistant to cyanide produced by Pseudomonas aeruginosa PA01 [23] . We therefore performed complementation testing between ye49 and the egl-9 null allele egl-9 ( sa307 ) and found that ye49/egl-9 ( sa307 ) animals are resistant to Cry21A PFT , indicating ye49 fails to complement egl-9 ( sa307 ) and most probably mutates the same gene ( Figure 1C ) . Furthermore , injection of an extrachromosomal array bearing the egl-9 promoter and coding region restored wild-type Cry21A susceptibility to ye49 animals ( Figure 1D ) . In addition , sequencing of egl-9 cDNA isolated from the ye49 mutant identified a point mutation ( W508-to-stop ) that upon translation is predicted to truncate the protein in the prolyl hydroxylase domain , thereby eliminating protein hydroxylase function ( Figure 1E ) . These results demonstrate that Cry21A PFT resistance phenotype associated with ye49 is due to loss of egl-9 function mutation . As predicted from this conclusion , feeding of double-stranded RNA to cause RNA interference ( RNAi ) results in animals resistant to Cry21A ( Figure 1F ) . The EGL-9 protein is a prolyl hydroxylase and functions in the C . elegans low oxygen response ( hypoxia ) pathway ( Figure 2A; [24] ) . The ability of cells to respond to hypoxia is mediated by a transcription factor called HIF-1α . Under normal oxygen ( normoxia ) conditions , HIF-1α ( called HIF-1 in C . elegans ) is hydroxylated by a prolyl hydroxylase ( EGL-9 in C . elegans or PHD in mammals ) that then increases HIF-1's affinity for von Hippel-Lindau tumor suppressor protein ( called VHL-1 in C . elegans ) , part of an E3 ubiquitin ligase complex . Association of HIF-1 with VHL-1 eventually leads to HIF proteasomal degradation . When EGL-9 is disabled by mutation , HIF-1 is stabilized at constitutively high levels even under normoxic ( normal oxygen ) conditions [25] . Since loss of EGL-9 function confers resistance to Cry21A PFT , we hypothesized that other elements of the hypoxia pathway might be important as well . We therefore performed quantitative dose-dependent mortality assays using null or putative null alleles of all the above elements of the hypoxia pathway . L4 staged animals from each genotype and wild-type N2 were placed in numerous doses of Cry21A PFT or Cry5B PFT and scored for viability after a few days ( Figures 2B and 2C ) . As predicted from the above studies , we find that animals lacking EGL-9 are quantitatively resistant to Cry PFTs . At doses from 1–16 µg/mL Cry21A and 10–80 µg/mL Cry5B PFTs , egl-9 ( sa307 ) and egl-9 ( ye49 ) animals are resistant to PFT-induced mortality relative to wild-type animals , with resistance strongest at higher PFT doses ( Figures 2B and 2C; Table 1 ) . For example , 7×–10× more egl-9 mutant animals are alive at 8 µg/mL Cry21A or 40 µg/mL Cry5B PFT than wild-type animals at the same doses ( P<0 . 001; Table 1; note , direct dose comparison between Cry21A and Cry5B toxicity is not possible since Cry21A assays are performed with Bt spore-crystal lysates and Cry5B assays with purified protein ) . Based on LC50 values ( concentration at which 50% of the animals are dead ) , loss of EGL-9 results in 3–5 fold resistance over wild-type animals to Cry21A or Cry5B PFTs ( Table 1 ) . Note , since all our mortality assays are carried out in liquid medium , resistance to the PFT cannot be attributed to improved avoidance behaviors . Cry21A PFT resistance was also confirmed using a quantitative brood size assay ( Figure S1 ) . We also found that vhl-1 ( ok161 ) mutant animals are resistant over a similarly wide range of Cry21A and Cry5B PFT doses ( Figures 2B and 2C; Table 1 ) . For example , 6 . 2× and 7 . 4× more vhl-1 mutant animals are alive at 8 µg/mL Cry21A and 40 µg/mL Cry5B , respectively , than wild-type animals . Based on LC50 values , vhl-1 mutant animals are 4× resistant to Cry5B PFT . We also tested rhy-1 ( ok161 ) mutant animals on Cry21A PFT . RHY-1 ( regulator of hypoxia-inducible factor ) antagonizes HIF-1 function by inhibiting expression of some HIF-1 target genes via a VHL-1 independent pathway [26] . Animals lacking RHY-1 are also resistant to Cry21A ( Table 1; Figure S2 ) . Based on LC50 values , animals lacking RHY-1 are 5 . 7× resistant to Cry21A PFT ( Table 1 ) . These data demonstrate that loss of function mutations in genes that normally antagonize HIF-1 function all result in resistance to Cry protein PFTs . In other words , stimulation of HIF-1 function via removal of HIF-1 inhibitory factors results in PFT resistance . To confirm that the resistance associated with egl-9 mutants was going through HIF-1 , we looked at the dose-dependent response of hif-1 ( ia04 ) and egl-9 ( sa307 ) hif-1 ( ia04 ) double mutant animals to Cry21A and Cry5B PFTs . When fed Cry21A , hif-1 ( ia04 ) animals have a response that is indistinguishable from wild-type animals ( Figure 2B; Table 1 ) . egl-9 ( sa307 ) hif-1 ( ia04 ) double mutant animals have a statistically identical response to Cry21A as wild-type animals at all doses except at 8 µg/mL ( Table 1 ) . Furthermore the LC50 values of hif-1 ( ia04 ) and egl-9 ( sa307 ) hif-1 ( ia04 ) animals on Cry21A PFT are statistically identical ( P>0 . 05 ) but both statistically different than that of egl-9 ( sa307 ) ( P<0 . 01; Table 1 ) . These results have been qualitatively confirmed using RNAi—whereas wild-type animals subject to egl-9 RNAi are resistant to Cry21A , hif-1 ( ia04 ) mutant animals subject to egl-9 RNAi are not ( Figure S3 ) . Similarly , whereas egl-9 ( sa307 ) mutants are resistant to Cry21A , this resistance is suppressed by RNAi of hif-1 . Thus , HIF-1 is required for Cry21A resistance mediated by loss of EGL-9 . The results with Cry5B PFT are similar to those of Cry21A ( Figure 2C; Table 1 ) in that HIF-1 is required for Cry5B resistance mediated by loss of EGL-9 ( i . e . , loss of HIF-1 suppresses Cry5B PFT resistance associated with loss of EGL-9 ) . There is , however , one striking difference between the hif-1 results with Cry21A and Cry5B . Both hif-1 ( ia04 ) and egl-9 ( sa307 ) hif-1 ( ia04 ) animals are hypersensitive to Cry5B PFT relative to wild-type animals . That is , animals lacking HIF-1 are more readily killed by Cry5B PFT than wild-type animals , especially at doses ∼5–10 µg/mL ( P<0 . 05; Figure 2C; Table 1 ) . Thus , hif-1 is required for intrinsic cellular defenses ( INCED ) [15] against Cry5B PFT . With regards to the different results with Cry5B and Cry21A , we speculate that perhaps Cry5B PFT intoxicates more potently than Cry21A and that , whereas increased HIF-1 activity is protective against all PFTs , loss of HIF-1 activity is more acutely felt when the stronger PFT is present . In the case of Cry21A , other INCED pathways are sufficient for full protection even in the absence of HIF-1 . Cry proteins are small-pore PFTs . To test whether or not the hypoxia pathway was more generally required for INCED against PFTs , we fed C . elegans two V . cholerae strains that differ primarily in their ability to produce another small-pore PFT , VCC . VCC is a virulence factor of V . cholerae and mutants lacking VCC are attenuated for pathogenesis in vivo , especially for strains lacking cholera toxin [27] , [28] . The strains we use are CVD109 ( VCC+ ) and CVD110 ( VCC− ) that are nearly isogenic ( except for the presence of cholera toxin B subunit in CVD110 , which should not matter since C . elegans lacks sialic acid that the B subunit binds to as part of its GM1 receptor [29] ) . Although both strains are pathogenic when fed to C . elegans , CVD109 ( VCC+ ) is more virulent than CVD110 ( VCC− ) , demonstrating that VCC is a virulence factor for C . elegans ( Figure 3A ) . Our results with hypoxia pathway mutants on CVD109 ( VCC+ ) and CVD110 ( VCC− ) are striking and parallel those with Cry PFTs . When feeding on CVD109 ( VCC+ ) , egl-9 ( sa307 ) animals are resistant relative to wild-type animals ( Figure 3A; Table 2; median survival 4 vs . 3 days respectively; P<0 . 001 ) . This resistance is dependent upon the presence of VCC since when feeding on CVD110 ( VCC− ) , egl-9 ( sa307 ) animals are not resistant ( Figure 3A; Table 2 ) . Similarly , hif-1 ( ia04 ) and egl-9 ( sa307 ) hif-1 ( ia04 ) animals are , as with Cry5B PFT , hypersensitive relative to wild-type animals on CVD109 ( VCC+ ) ( median survival of 2 , 1 , and 3 days respectively; P<0 . 0001; Figure 3B; Table 2 ) . This hypersensitivity is dependent upon the presence of VCC since these mutants are not hypersensitive when feeding on CVD110 ( VCC− ) ( Figure 3B; Table 2 ) . It is interesting to note that egl-9 ( sa307 ) mutant animals are hypersensitive compared to wild-type animals to CVD110 ( VCC− ) strain ( median survival of 4 and 6 days respectively; P<0 . 0001; Figure 3A; Table 2 ) . We speculate that while activation of the hypoxia pathway ( in an egl-9 mutant or otherwise ) protects the animals against VCC and PFTs ( hence egl-9 mutants are resistant to the VCC+ strain ) , activation of the hypoxia pathway may make the animals more susceptible to other V . cholerae virulence factors . The relative contribution to these responses ( protection versus susceptibility ) is dependent upon which virulence factors are present and their relative contribution to virulence . In the VCC+ strain , the PFT has important function . Hence , the protective role of pathway activation can be discerned . In the VCC− strain , the PFT defense is no longer needed . Hence , the susceptible role can be discerned . It is this give-and-take interaction between the host and virulence factors that could partly explain why constitutive mutation in egl-9 is not selected in the wild . In any event , taken together , our Cry PFT and VCC data demonstrate that stabilization of HIF-1 results in resistance to VCC PFT whereas loss of HIF-1 results in hypersensitivity to VCC PFT . Because loss of EGL-9 results in resistance to PFTs ( here ) and cyanide [23] , [30] , we hypothesized that egl-9 mutant animals might show resistance to other stressors as well . We found that , relative to wild-type animals , animals lacking EGL-9 are resistant to killing by 1 ) the pathogen Pseudomonas aeruginosa PA14; 2 ) heat stress; and 3 ) oxidative stress ( Figure 4A , Table 2; Figures 4B and 4C ) . Since correlation between stress response and lifespan had previously been reported , such as in the daf-2 mutant [31] , [32] , we tested whether loss of EGL-9 had an effect on longevity . Indeed , egl-9 ( ye49 ) and egl-9 ( sa307 ) mutant animals live longer than N2 wild-type when feeding on the standard E . coli strain ( Figure 4D , Table 2 ) . To study the relationship between the hypoxia response pathway and resistance to stresses in more detail , we asked if the resistance to these different stresses via loss of EGL-9 was , as for resistance to PFTs , mediated through HIF-1 . Unexpectedly , we found that hif-1 ( ia04 ) loss-of-function mutant animals as well as egl-9 ( sa307 ) hif-1 ( ia04 ) mutant animals are resistant to P . aeruginosa PA14 infection , heat stress , and oxidative stress ( Figure 4E , Table 2; Figures 4F and 4G ) . Both mutant strains are also long lived ( Figure 4H; Table 2 ) . Thus , in the case of these stresses , but unlike that of PFT response , loss of either EGL-9 , HIF-1 , or both results in stress resistance . We speculate that , in the case of these other stresses , hydroxylation of HIF-1 by EGL-9 may result in its activation prior to degradation . Similar results have been previously reported in that mutation of either hif-1 or egl-9 results in C . elegans resistant to pathogenic E . coli [33] . With regards to lifespan , published studies are contradictory but there is at least one published report with egl-9 mutants long lived and two with hif-1 long-lived [34] , [35] , [36] . In any event , our results demonstrate that role of the hypoxia pathway in PFT INCED is separable from that of other stress responses . Bt Cry PFTs attack intestinal cells [21] , [22] , [37] . It is possible that the hypoxia defense pathway functions within the cells targeted by the PFTs or that the hypoxia pathway is functioning cell non-autonomously . To address this question , we expressed egl-9 under the control of various promoters including the intestinal specific cpr-1 promoter [21] , [22] , [38] and the unc-31 promoter , which is expressed in all neurons and in secretory cells of the somatic gonad [39] . We find that expression of wild-type EGL-9 under the cpr-1 promoter in the intestinal cells of egl-9 ( sa307 ) animals ( Figure 5 ) , but not under the unc-31 promoter in the neuronal or secretory cells ( not shown ) , is sufficient to rescue the egl-9 ( sa307 ) Cry21A resistance phenotype . Control animals in which green-fluorescent protein ( GFP ) was expressed from the cpr-1 promoter did not result in rescue . Quantitative mortality assays using two independent lines of cpr-1::egl-9-transformed egl-9 ( sa307 ) mutant animals confirm that intestinal-specific expression of EGL-9 rescues Cry21A PFT resistance to a level statistically indistinguishable from N2 wild-type ( not shown ) . These data are consistent with the hypoxia pathway acting to directly counteract the effects of PFTs and not , for example , providing protection via altered behavior . To address how the hypoxia pathway might function in protection against PFTs , we sought in two ways to find functional downstream effectors of the pathway . First , we compared known functional targets of the hypoxia pathway in C . elegans and asked if any of these are involved in PFT defenses . One pathway immediately surfaced , the unfolded protein response or UPR [40] . It has been recently reported that the hypoxia pathway genetically functions upstream of the XBP-1 arm of the UPR with regards to longevity in C . elegans [35] . Furthermore , we have previously shown that the XBP-1 is required for PFT INCED since loss of XBP-1 leads to animals that are hypersensitive to Cry5B PFT [15] . These data suggest that the XBP-1 arm of the UPR is one downstream target of the hypoxia PFT INCED . To test this suggestion , we examined whether or not the hypoxia pathway regulates activation of the XBP-1 UPR pathway . Activation of the XBP-1 UPR pathway can readily be discerned by examining xbp-1 mRNA , which is spliced upon activation of the pathway [41] . We indeed find that activation of the hypoxia pathway results in activation of the UPR as seen by a 1 . 4 fold increase in spliced xbp-1 levels in egl-9 mutant animals ( P<0 . 001; see Materials and Methods ) . Thus , one functional downstream effector of the hypoxia pathway for PFT defenses is the XBP-1 UPR . We conversely asked if any of the genes known to be involved in PFT INCED are known to be important for the hypoxia pathway . From over 100 PFT INCED genes we have identified in our lab , we found one and only one currently known to be regulated by the hypoxia pathway , nhr-57 . nhr-57 was initially identified as part of the hypoxia pathway by the fact that its expression is positively regulated by hif-1 and negatively regulated by egl-9 and vhl-1 [42] , [43] . In fact , nhr-57 transcriptional activation is considered the most reliable marker for activated HIF-1 function in C . elegans [26] . We confirmed using quantitative PCR that in egl-9 mutant animals , nhr-57 transcripts are induced 15 fold and that this increase is completely dependent upon HIF-1 ( data not shown ) . However , to date no functional role of nhr-57 for any HIF-1-regulated pathway has been shown . We find that knock down of nhr-57 results in animals slightly but statistically hypersensitive to Cry5B PFT ( e . g . , 21% reduction in viability for nhr-57 RNAi at 20 µg/mL Cry5B PFT versus vector-only RNAi control , P = 0 . 02; n = 90; see Materials and Methods ) and therefore defective in PFT INCED . More impressively , we find that knock down of nhr-57 completely suppresses the resistance to Cry21A PFT associated with loss of EGL-9 ( Figure 6 ) . Taken together , these results indicate that the nuclear receptor nhr-57 is a second functional downstream effector of the hypoxia PFT defense pathway . Although the above data demonstrate the hypoxia pathway is important for PFT INCED , they do not directly address whether the defense against PFTs is related to a low oxygen response or to some other function of the HIF-1 pathway . We therefore examined whether the hypoxia pathway itself is activated by PFTs using nhr-57 expression , the canonical marker for HIF-1 pathway activation by low oxygen in C . elegans ( see above ) . We find that 4 and 8 hours of treatment with PFT significantly induces nhr-57 expression 5 . 3 and 3 . 6 fold respectively ( Figure 7A ) . Shorter treatments with PFT do not . Thus , PFT induces the hypoxia pathway . If a low oxygen response is involved in responding to PFTs , then one might predict that exposure to low oxygen might confer protection against PFT attack since the low oxygen environment might strongly and rapidly induce the correct protective response . We therefore exposed C . elegans hermaphrodites to low ( 2% ) oxygen levels minus or plus the presence of E . coli-expressing Cry5B PFT . We find that low oxygen is indeed protective against PFT intoxication in that animals exposed to PFT in a low oxygen environment for 24 hours are significantly healthier than animals exposed to PFT in normoxia ( Figure 7B ) . Similar results were obtained for animals exposed to a low oxygen environment for three days ( Figure S4 ) . In contrast and as expected , hif-1 ( ia04 ) mutant animals exposed to Cry5B PFT do not get any protection when placed in a hypoxic environment ( Figure 7B ) , confirming that the protective effect of hypoxia against PFT is due to activation of the HIF-1 pathway .
Our results demonstrate that the hypoxia pathway protects C . elegans against PFTs , whether Bt Cry protein PFTs or a PFT used by a mammalian pathogen , V . cholerae VCC . We find that activation of HIF-1 pathway by removal of any of EGL-9/PHD , VHL-1 , or RHY-1 , makes C . elegans more resistant to PFTs than they normally are . This resistance is completely abrogated upon loss of HIF-1 , which can additionally result in animals hypersensitive in PFTs . Resistance to PFTs functions in the cells directly targeted by PFTs and is not associated with other hypoxia-mediated stress resistance phenotypes . Furthermore , exposure to PFT induces transcriptional activation of the HIF-1 low oxygen pathway , and exposure of animals to low oxygen protects animals against PFT intoxication , through a HIF-1-dependent mechanism . A schematic summarizing our findings here is in Figure 7C . Consistent with our finding that activation of the HIF-1 pathway is protective against PFTs , it has been shown that expression of the HIF-1α protein is increased in human airway cells by S . aureus supernatants , of which α-toxin is a major constituent [44] . The simplest interpretation of our data is that PFT intoxication is associated with low oxygen in cells , and that the hypoxia pathway is therefore needed to protect the cells against this condition . Alternatively , although less parsimoniously , it is possible that both hypoxia and PFTs trigger the same set of HIF-1 downstream mediators that are protective against both assaults but that are not otherwise linked by the presence of low oxygen . Two downstream effectors of the hypoxia PFT INCED pathway are the UPR and nhr-57 . The fact that nhr-57 is involved in hypoxia PFT INCED suggests that multiple transcriptional responses are key to mounting an effective defense against PFTs . The link between the XBP-1 UPR , hypoxia , and PFTs is intriguing . It has already been shown in mammalian cells that hypoxia induces activation of the XBP-1 UPR as detected by an up-regulation in xbp-1 mRNA splicing by low oxygen [45] . Furthermore , it has been shown that XBP-1 protects cells against hypoxia-induced apoptosis [45] . Therefore , we speculate that one role of the hypoxia pathway in PFT INCED is to induce an XBP-1-linked protective response against PFT/low oxygen-mediated apoptosis . Given that the p38 MAPK pathway is also linked to PFT defenses and the UPR , it will be interesting to explore further links between hypoxia , the UPR , p38 and apoptotic pathways in response to PFTs . Although this report is not the first of hypoxia pathway involvement in immunity , it is the first showing a link between hypoxia and protection of cells that are being directly attacked by a virulence factor . Control of the metabolic shift to glycolysis by HIF-1α has been shown to play an important role in myeloid cell-mediated inflammatory response [46] . Furthermore , it has been shown that bacteria increase HIF-1α protein expression and stimulate HIF-1α transcriptional activity in macrophages , regulating the expression of immune effectors molecules , including antimicrobial peptides , nitric oxide and tumor necrosis factor-α [47] . Our results point to a new and different role of the hypoxia pathway , namely in providing autonomous protection of epithelial cells against PFTs . To our knowledge , these results are the first to demonstrate that an intracellular pathway can be altered to promote general resistance to PFTs . Although a few receptor mutants that confer resistance to PFTs have been previously identified , these do not confer general resistance . A logical extension of our findings is that significant therapeutic benefit against a wide range of bacterial pathogens such as S . aureus , Streptococci , Clostrida , V . cholerae ( all of which use PFTs as virulence factors ) could be achieved by up-regulation of HIF-1 and/or by hypoxia . The identification of the hypoxia pathway as an important PFT INCED pathway thus unexpectedly provides a novel and potentially powerful means of protecting against the single most common mode by which bacterial pathogens attack us .
Strains were maintained at 20°C under standard conditions [48] . The wild-type strain for this study is N2 Bristol [48] . The strains egl-9 ( sa307 ) , hif-1 ( ia04 ) , egl-9 ( sa307 ) hif-1 ( ia04 ) , vhl-1 ( ok161 ) , rhy-1 ( ok1402 ) , the Hawaiian strain CB4856 and HT1593 [unc-119 ( ed3 ) ] were obtained from the Caenorhabditis Genetic Center ( CGC ) . All strains were either previously outcrossed or outcrossed here at least six times ( e . g . , egl-9 ( ye49 ) , rhy-1 ( ok1402 ) ) . egl-9 ( sa307 ) is a null allele of egl-9 that carries an internal 243-bp deletion removing part of exons 5 and 6 [23] . hif-1 ( ia04 ) allele removes exons 2 , 3 and 4 of hif-1 , including the DNA binding domain , and is believed to be a null allele [49] . The vhl-1 ( ok161 ) allele removes exons 1 and 2 of vhl-1 and is believed to be a null allele [25] . The rhy-1 ( ok1402 ) allele deletes exons 2 , 3 and 4 of rhy-1 and is also believed to be null [26] . Images were acquired using an Olympus SZ60 dissecting microscope and a Canon PowerShot A620 digital camera . For production in Bt , the cry21A gene was cloned under 700 bp of the cry6A promoter region and subcloned into the Bt/E . coli shuttle vector pHT3101 . The plasmid was transformed into a nontoxic host Bt strain ( 4D22 ) . Cry21A SCLs were prepared using standard procedures [50] and the concentration was measured relative to BSA standards on protein gels . Mutagenesis and selection of Cry21A resistance mutants was carried out as described for Cry5B [20] except Cry21A SCLs were used to a final concentration of 0 . 25 µg/mL Cry21A . The 68 , 000 F2 animals were taken from a larger population of 1 , 300 , 000 F2 animals that came from 240 , 000 mutagenized F1 animals . Animals were incubated for 72 hours at 20°C and scored for overall health , including color , size , movement and brood size . Clonal lines were established from candidates and retested . Complementation tests were performed by testing F1 progeny from the cross between egl-9 ( sa307 ) males and dpy-17 ( e164 ) ;ye49 hermaphrodites . As a control , cross-progeny from egl-9 ( sa307 ) males into dpy-17 ( e164 ) and from N2 males into dpy-17 ( e164 ) ;ye49 were also tested . ye49 was mapped between dpy-11 ( e224 ) and unc-76 ( e911 ) using standard three-factor mapping . A dpy-11 ( e224 ) ye49 unc-76 ( e911 ) triple mutant was then made in order to perform single nucleotide polymorphism mapping with the Hawaiian strain ( CB4856 ) [51] . Genomic DNA and cDNA prepared from egl-9 ( ye49 ) animals were used to sequence the egl-9 gene . For transformation rescue , a 13 . 4kb-PCR fragment covering from 3kb upstream to 2kb downstream of egl-9 transcript was amplified with primers GAGCAACTCGTGGGTTTGTT and CTTCCAGAGGCGTTGTCTTC using the LongAmp Taq ( Biolabs ) from N2 genomic DNA and injected in egl-9 ( ye49 ) worms as described [52] . For tissue-specific rescue , egl-9 rescuing plasmids were constructed by PCR amplification of unc-31 and cpr-1 promoters and then fused to egl-9 and gfp open reading frames using the Multisite Gateway cloning system ( Invitrogen ) and pCG150 ( containing unc-119 rescuing fragment ) [53] . The constructs were verified by sequencing and integrated into HY0843 [unc-119 ( ed3 ) ;egl-9 ( sa307 ) ] by ballistic bombardment [54] with a PDS-1000/He Biolistic Particle Delivery System ( Bio-Rad , Hercules , CA ) . Two independent lines of each transgenic strain were examined . For Cry21A E . coli toxin assays , we used E . coli JM103 with pQE9 empty vector or a cry21A gene insert under control of the lacZ promoter [11] . Since Cry21A is expressed at very high levels by E . coli [11] and too potent for scoring for resistance , we diluted the toxin-expressing bacteria with non-toxin-expressing bacteria at a ratio of 1∶40 for all tests in this study , similar to that previously described for Cry5B studies [13] , [15] . Dose-dependent mortality assays with purified Cry5B were performed as described [52]; hermaphrodite viability was scored after 6 days at 25°C . Cry21A SCLs were used for quantitative mortality assays as described above except hermaphrodite viability was scored after 72 hours at 20°C . Each assay was set up with triplicate wells for each concentration of Cry toxin , and each experiment was performed in at least three independent trials . Typically 180 worms were scored for each concentration . V . cholerae lifespan assay was performed as described [55] except the overnight culture was spread on Brain Heart Infusion ( BHI ) plates . CVD109 Δ ( ctxAB zot ace ) and CVD110 Δ ( ctxAB zot ace ) hlyA:: ( ctxB mer ) Hgr strains , derived from V . cholerae El Tor E7946 , were used [56] . The experiment was performed three times with approximately 50 worms per strain at room temperature ( 22°C ) . P . aeruginosa lifespan assays were performed on slow-killing plates as described [15] . Heat shock assays were performed as described [57] . For the oxidative stress analysis , synchronized young adults were exposed to 7 . 5 mM t-butyl hydrogen peroxide as described [58] and were observed after 6 hours . Life-span assays were initiated by allowing adult hermaphrodites to lay eggs on NG plates spread with OP50 . When these eggs hatched and the nematodes reached the L4 stage they were transferred to fresh NG plates with OP50 supplemented with 25 µM 5-fluorodeoxyuridine ( FUDR ) to prevent eggs from hatching . The nematodes were scored for live/dead every 48 hours by tapping the nose at least three times ( no movement for all taps was scored as dead ) . For RNAi tests , adult hermaphrodites were allowed to lay eggs on NG plates containing 100 µM Isopropyl β-D-1-thiogalactopyranoside ( IPTG ) and 50 µg/mL ampicillin spread with E . coli strain HT115 expressing double-stranded ( ds ) RNA ( from the Ahringer library [59] ) for 8 hours and then removed . The eggs were allowed to develop into L4 larvae on RNAi plates at 20°C . L4 hermaphrodites ( ten per genotype or line ) were picked onto toxin plates spread with 100 µl of a mixture of E . coli strain HT115 expressing the same dsRNA and HT115 harboring cry21A-expressing vector at the ratio 40∶1 . For no toxin control plates , 100 µl of HT115 with dsRNA was spread . nhr-57 ( RNAi ) testing on Cry5B PFT was performed slightly differently ( Kao et al . , manuscript in preparation ) . Briefly , rrf-3 ( pk1426 ) animals were fed E . coli-expressing dsRNA in liquid media with 1mM IPTG at 25°C for ∼30 h . 20 µg/ml of Cry5B or 20 mM HEPES control were then added , as well as 200 µM FUdR . Hermaphrodites viability was scored after 6 days at 25°C . ( As this assay is set up differently , direct comparison with dose-dependent mortality assays presented in Table 1 and associated Figures is not possible ) . To test worms under hypoxia , L4 wild type and hif-1 ( ia04 ) mutant animals were pipetted onto toxin plates spread with 30 µl of a mixture of E . coli OP50 strains expressing or not Cry5B at the ratio 1∶1 . Plates were placed immediately in a 2% O2 chamber for 24 hours , while control plates were placed in room air . Images were taken with an Olympus BX60 microscope as described [15] . Real time PCR was performed as described [15] . To determine the levels of spliced xbp-1 mRNA , we used primers xbp-1_sqf2 GCATGCATCTACCAGAACGTC and xbp-1_sqr2 GTTCCCACTGCTGATTCAAAG to amplify cDNA from wild-type and egl-9 ( sa307 ) animals . The forward primer xbp-1_sqf2 anneals to exon 1 and the reverse primer xbp-1_sqr2 anneals to the exon1-exon 2 junction sequence produced when intron 1 is spliced out . The experiment was carried out using two independent sets of cDNA and two repeats within each set . Primers TTATCGAGTTTCTCGCATTGG and AAGTCTGCAATCACGCTCTGT were used to quantify expression of nhr-57 . Induction of expression of nhr-57 by Cry5B was tested in glp-4 ( bn2 ) animals treated for 1 , 2 , 4 and 8 hours on E . coli OP50 strains expressing Cry5B or not . The experiment was carried out using three independent sets of cDNA . Normalization in all cases was to eft-2 transcript levels . LC50 values were determined by PROBIT analysis [60] . Mortality assays were plotted using GraphPad Prism 5 . 0 ( San Diego ) . Statistical analysis between two values was compared with a paired t-test . Statistical analysis among three or more values of one independent variable was done with matched one-way ANOVA with Tukey's method and of more than two independent variables by two-way ANOVA with the Bonferroni post test . For lifespan analysis , survival fractions were calculated using the Kaplan-Meier method and survival curves compared using the logrank test . Statistical significance was set at P<0 . 05 . | Bacteria make many different protein toxins to attack our cells and immune system in order to infect . Amongst them , pore-forming toxins ( PFTs ) , which punch holes in the protective plasma membrane that surrounds cells , are by far the most abundant and constitute important virulence factors . Since the integrity of the plasma membrane is fundamental to maintaining the normal intracellular environment , the breaching of the plasma membrane by PFTs results in many and dramatic intracellular responses . However , we know little about the relevance of these responses to cell survival or cell intoxication . Here , using the only genetic system for studying pore-forming toxin effects in a whole animal , we show that the same response that protects cells against low oxygen stress unexpectedly also protects cells against pore-forming toxins . Mutations in the animal that hyper-activate the low oxygen response actually make animals resistant to pore-forming toxin attack , whereas mutations that inactivate the low oxygen response make animals more susceptible . Furthermore , a low oxygen environment itself is protective against pore-forming toxins . These data show a new and powerful connection between low oxygen responses and defense against the single most common mode of bacterial attack . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/animal",
"genetics",
"infectious",
"diseases",
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses",
"cell",
"biology/cell",
"signaling",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system",
"immunology/immune"... | 2009 | Hypoxia and the Hypoxic Response Pathway Protect against Pore-Forming Toxins in C. elegans |
Japanese encephalitis ( JE ) is a mosquito-borne disease that is associated with considerable morbidity and mortality in many Asian countries . The objective of this study was to describe the impact of the JE immunization program using SA 14-14-2 JE vaccine implemented in Nepal during 2006 through 2011 . A previous assessment after the initial program implementation phase described a significantly lower post-campaign JE incidence compared to expected incidence; however , the previous evaluation had limited post-campaign data for some districts . JE and acute encephalitis syndrome ( AES ) data gathered through Nepal’s routine surveillance system from 2004 through 2014 were analyzed to assess the impact of the JE immunization program implemented in 31 districts . Expected incidence rates were determined by calculating the incidence of cases per 100 , 000 person-years in each district before the vaccination campaigns . This rate was applied to the relevant population after the vaccination campaigns , which provided the expected number of cases had the campaign not occurred . The observed incidence rate was the number of reported cases per 100 , 000 person-years post-campaign . Expected and observed JE and AES cases and incidence rates were compared . The post-campaign JE incidence rate of 0 . 7 cases per 100 , 000 was 78% ( 95% CI 76%–79% ) lower than expected had no campaign occurred and an estimated 3 , 011 ( 95% CI 2 , 941–3 , 057 ) JE cases were prevented . The post-vaccination AES incidence of 5 . 5 cases per 100 , 000 was 59% ( 58%–60% ) lower than the expected and an estimated 9 , 497 ( 95% CI 9 , 268–9 , 584 ) AES cases were prevented . This analysis strengthens previous findings of the substantial impact of Nepal’s JE immunization program using SA 14-14-2 JE vaccine .
Japanese encephalitis ( JE ) is a mosquito-borne disease that is associated with considerable morbidity and mortality in many Asian countries [1] . Approximately 20–30% of JE cases are fatal and 30–50% of survivors have neuropsychiatric sequelae [1 , 2] . There is no specific treatment for JE , but the disease is preventable by vaccination . During the past decade , there has been a substantial increase in the availability of improved JE vaccines , including the live-attenuated SA 14-14-2 JE vaccine [3] . The SA 14-14-2 vaccine has been found to be both safe and effective with over 96% protective effectiveness five years after administration of a single dose [4] . WHO recommends the use of JE vaccine in areas where JE is a public health priority , and has highlighted the need for good quality disease surveillance data to monitor vaccine impact [5] . JE was first recognized as a public health problem in Nepal in the mid-1970s [6 , 7] . In 2006 , the Ministry of Health and Population ( MOHP ) in Nepal commenced an immunization program using SA 14-14-2 JE vaccine , and by 2009 , the program had been implemented in 23 districts with the highest JE disease burden . An impact assessment in 2010 found that the post-campaign JE incidence rate was 72% lower than expected if no campaigns had occurred , and the clinical acute encephalitis syndrome ( AES ) incidence was 58% lower than expected [8] . However , limited post-introduction data in some districts might have impacted the accuracy of the assessment of the full impact of the program . In 2011 , Nepal MOHP expanded the immunization program to an additional eight districts . By 2014 , JE and AES surveillance data had been gathered for more than 3 years in all districts where the program had been implemented . In this assessment , we included JE and AES surveillance data from 2004 through 2014 , thereby adding surveillance data for 2010–2014 for the 23 districts originally included in the impact analysis and data from 2004–2014 for the additional eight districts . We reviewed and analyzed these data to assess the impact the JE immunization program in the 31 districts .
Nepal is divided into three ecological zones from north to south; the Himalayan mountain region along the northern border with China , the hill region in the middle , and the Terai along the southern border with India . More than half of Nepal’s population of approximately 28 million live in the 24 districts of the Terai region . This area has conditions highly favorable for JE virus transmission; historically , more than 90% of JE cases in Nepal have been reported from this region [9 , 10] . In 1995 , JE virus transmission was confirmed in Kathmandu Valley and endemicity was subsequently documented in other hill and mountain districts [11 , 12] . In this analysis , as in the previous assessment , the campaign districts were categorized according to JE risk level [8] . Four western Terai districts ( Kailali , Bardiya , Banke , and Dang ) were classified as high risk; districts with lower incidence rates but recurrent seasonal JE virus transmission were classified as moderate risk . From 2006 through 2011 , mass immunization campaigns were conducted in 31 ( 41% ) of Nepal’s 75 administrative districts , starting with Terai districts with the highest JE disease burden . By the end of 2011 , campaigns had been implemented in all 24 Terai districts and seven ( 20% ) of 35 hill districts . In 20 districts , campaigns targeted all persons aged ≥1 year; in the other 11 districts , only children aged 1–15 years were vaccinated . The median reported campaign coverage rate for the 31 districts was 90% ( range: 59%–115% ) [13]; coverage rates higher than 100% might have been due to residents of other districts being vaccinated and included in counts or underestimates of a district’s population . Routine JE immunization for children aged 12–23 months was introduced within 3 years of the mass campaign in each district . A single dose of SA 14-14-2 JE vaccine was used in campaigns and in the routine program . Routine surveillance for AES cases in Nepal began in 1978 . In 2004 , AES and JE surveillance programs were strengthened , including the designation of 45 medical facilities as sentinel sites , initiation of enhanced case-based surveillance using a standardized case definition , and improved access to JE laboratory testing [10] . The initial 45 sentinel sites consisted of 34 sites in 20 Terai districts and 11 sites in 5 hill districts [8] . The AES/JE surveillance system was expanded to additional sites in subsequent years . AES cases identified at sentinel sites are reported to the Programme for Immunization Preventable Diseases at the World Health Organization ( WHO ) in Nepal . Cerebrospinal fluid ( CSF ) and serum specimens are collected when possible and tested at the National Public Health Laboratory or B . P . Koirala Institute of Health Sciences laboratory using a JE immunoglobulin ( Ig ) M antibody capture enzyme-linked immunosorbent assay ( MAC-ELISA ) [8] . Over the surveillance period , a variety of assays were used , including commercially available kits and laboratory-developed assays . Epidemiological and laboratory data were collected in a MOHP/WHO database maintained by WHO Nepal . Nepal uses WHO-recommended case definitions . An AES case is a person of any age , at any time of the year with the acute onset of fever and a change in mental status and/or new onset of seizures , excluding simple febrile seizures [14] . All cases that meet this definition are included in clinical AES surveillance , regardless of whether laboratory testing is performed . An AES case with JE virus-specific IgM antibody detected in CSF or serum is considered a JE case [14] . JE and AES surveillance data for this analysis were included from the 45 reporting sites established in 2004 , also used for the previous impact analysis , and from two sites established in hill districts in 2005 . We included these two additional reporting sites because more than one third of AES cases from two of the new districts presented to these sites ( and not to one of the 45 sites used in the previous analysis ) . The same basic analysis methodology was used as documented in the previous assessment [8] . Briefly , JE and AES expected incidence rates were determined by calculating the incidence per 100 , 000 person-years in each district or age group before the vaccination campaign . This rate was applied to the relevant population after the vaccination campaign to calculate the expected number of cases had the campaign not occurred . The observed incidence rate was the number of reported cases per 100 , 000 person-years post-campaign . The differences between expected and observed incidence rates and exact binomial confidence intervals for the differences were calculated . The cut-off dates between the “pre” and “post” vaccination campaign periods for each district were determined by selecting the mid-point of the vaccination campaign and adding 2 weeks to allow for development of immunity after vaccination . Campaigns were typically completed over a period of 2 weeks to 2 months . In three districts , partial campaigns were conducted in two consecutive years . To provide a conservative estimate of impact in these districts , the cut-off date definition was applied in the first year of the campaign . In short , cases that occurred from 2004 until 2 weeks following the mid-point of the vaccination campaign were used to calculate expected incidence rates , and those that occurred >2 weeks after the mid-point of the campaign until the end of 2014 were used to calculate observed rates . The annual population data used for incidence calculations were obtained from the Nepal Department of Health Services’ Health Management Information System . These data are the official estimates used by the Nepal MOHP . SAS version 9 . 3 ( SAS Institute , Cary , NC ) and R version 3 . 1 . 2 were used for data analysis . The protocol and analytical approach for this project were reviewed at the U . S . Centers for Disease Control and Prevention and the Nepal MOHP , respectively , and were considered to be program evaluation . Therefore , institutional review board review was not required . For epidemiological analysis , the data were anonymized .
In the 31 districts , the observed post-campaign JE incidence rate was 0 . 7 cases per 100 , 000 person-years , 78% lower ( 95%CI 76%–79% ) than the expected incidence of 3 . 3 cases per 100 , 000 ( Table 1 ) . The vaccination campaigns prevented an estimated 3 , 011 ( 95%CI 2 , 941–3 , 057 ) JE cases . The median difference between observed and expected incidence among the 31 districts was 71% lower ( range: 100% lower–0% change ) ( Fig 1 ) . The difference was significantly lower in 28 ( 90% ) districts . The estimated impact was similar among all age groups ( Table 2 ) . The greatest impact was seen in the four high-risk Terai districts where the observed incidence rate was 89% lower ( 95%CI 87%–90% ) than expected , and an estimated 1 , 955 JE cases were prevented . In the moderate risk areas , the observed post-vaccination incidence was 62% lower ( 95%CI 59%-65% ) in the 20 Terai districts and 69% lower ( 95%CI 63%–75% ) in the seven hill districts . An estimated 819 and 237 JE cases were prevented in those regions , respectively ( Table 1 ) . The observed post-campaign AES incidence rate of 5 . 5 cases per 100 , 000 person-years was 59% lower ( 95%CI 58%–60% ) than the expected incidence of 13 . 7 cases per 100 , 000 ( Table 3 ) . An estimated 9 , 497 ( 95%CI 9 , 268–9 , 584 ) AES cases were prevented by the vaccination campaigns . The median difference between observed and expected incidence was 41% lower ( range: 96% lower–95% higher ) ; the difference was significantly lower in 22 ( 71% ) districts ( Fig 2 ) . Among the six districts where observed AES incidence was higher than expected , five were adjacent districts in the central Terai and one was in the central Hill region . As with JE , the greatest impact was in the four high-risk Terai districts where the observed incidence was 88% lower ( 95%CI 88%–89% ) than expected . In the moderate risk areas , the observed incidence was 28% lower ( 95%CI 25%–30% ) in the 20 Terai districts and 42% lower ( 95%CI 39%–45% ) in the seven hill districts .
The findings and conclusions in this report are the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention . | In 2006 , the Ministry of Health and Population in Nepal commenced a Japanese encephalitis ( JE ) immunization program using SA 14-14-2 JE vaccine , with mass campaigns conducted in selected districts , followed by introduction of JE vaccine into the routine childhood immunization program . JE and acute encephalitis syndrome data gathered through Nepal’s routine surveillance system from 2004 through 2014 were analyzed to assess the impact of this immunization program . Expected and observed JE and acute encephalitis syndrome cases and incidence rates were compared . Considerable impact on JE incidence was demonstrated and the results also suggested that a large proportion of acute encephalitis syndrome cases without laboratory confirmation are due to JE . The results support the belief that a JE immunization program will result in sizable reductions in the incidence of both laboratory-confirmed JE and clinical acute encephalitis syndrome cases . JE is a severe disease , and the program’s impact likely extended to reduction of rates of JE-related mortality and long-term disability . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"death",
"rates",
"medicine",
"and",
"health",
"sciences",
"demography",
"immunology",
"geographical",
"locations",
"tropical",
"diseases",
"encephalitis",
"vaccines",
"preventive",
"medicine",
"age",
"groups",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",... | 2017 | Updated estimation of the impact of a Japanese encephalitis immunization program with live, attenuated SA 14-14-2 vaccine in Nepal |
A PCR-enzyme-linked immunosorbent assay ( PCR-ELISA ) was developed to overcome the need for sensitive techniques for the efficient diagnosis of Schistosoma infection in endemic settings with low parasitic burden . This system amplifies a 121-base pair tandem repeat DNA sequence , immobilizes the resultant 5′ biotinylated product on streptavidin-coated strip-well microplates and uses anti-fluorescein antibodies conjugated to horseradish peroxidase to detect the hybridized fluorescein-labeled oligonucleotide probe . The detection limit of the Schistosoma PCR-ELISA system was determined to be 1 . 3 fg of S . mansoni genomic DNA ( less than the amount found in a single cell ) and estimated to be 0 . 15 S . mansoni eggs per gram of feces ( fractions of an egg ) . The system showed good precision and genus specificity since the DNA target was found in seven Schistosoma DNA samples: S . mansoni , S . haematobium , S . bovis , S . intercalatum , S . japonicum , S . magrebowiei and S . rhodaini . By evaluating 206 patients living in an endemic area in Brazil , the prevalence of S . mansoni infection was determined to be 18% by examining 12 Kato-Katz slides ( 41 . 7 mg/smear , 500 mg total ) of a single fecal sample from each person , while the Schistosoma PCR-ELISA identified a 30% rate of infection using 500-mg of the same fecal sample . When considering the Kato-Katz method as the reference test , artificial sensitivity and specificity rates of the PCR-ELISA system were 97 . 4% and 85 . 1% , respectively . The potential for estimating parasitic load by DNA detection in feces was assessed by comparing absorbance values and eggs per gram of feces , with a Spearman correlation coefficient of 0 . 700 ( P<0 . 0001 ) . This study reports the development and field evaluation of a sensitive Schistosoma PCR-ELISA , a system that may serve as an alternative for diagnosing Schistosoma infection .
Schistosomiasis affects 200 million people and about 779 million people live in endemic areas in the Middle East , South America , Caribbean , Southeast Asia and particularly sub-Saharan Africa [1] . Population- and treatment-based control programs have been successful in reducing the intensity of infection and severe morbidities associated with schistosomiasis; however , transmission remains active in highly endemic areas , and recurring low-level reinfection is likely to be associated with subtle but persistent morbidities such as anemia , malnutrition and diminished performance status [2]–[4] . In the presence of these conditions , the assessment of infection becomes less reliable since the currently used diagnostic methods are not sufficiently sensitive to accurately determine the prevalence of schistosomiasis or parasite burden in order to eventually achieve elimination of the disease [5] , [6] . Microscopic demonstration of the parasite's eggs in feces or urine remains the most wide-spread tool for schistosomiasis diagnosis . The Kato-Katz technique [7] is currently the most used method for fecal examination because it is quantitative , relatively inexpensive and simple . A significant increase in the sensitivity of the method is gained by microscopic examination of multiple samples [8] , [9] , but this is a limiting procedure for field work . To overcome the current limitations with respect to diagnosis , the simultaneous use of different diagnostic methods , such as antibody detection followed by stool examination of seropositive individuals , has been applied to monitor the human population and to identify the small number of infected people once morbidity control is achieved [6] . However , because antibody detection methods often cannot distinguish between current and past infection and may also present a high level of crossreactivity , molecular tools should be considered despite their higher cost and the requirement for special laboratory equipment [10] . Hamburger et al . [11] described a 121-base pair tandem repeat DNA sequence present in 12% of Schistosoma mansoni genome . This sequence has been successfully used in PCR-based approaches for the detection of the parasite in snails [12] , monitoring of cercariae in water bodies [13] and diagnosis of human infection using fecal or serum samples [14] and , more recently , plasma samples [15] . In a population study , the prevalence of S . mansoni infection was determined to be 31% when three fecal samples were examined using the Kato-Katz technique , but the prevalence rose to 38% when the PCR technique developed by Pontes et al . [14] was employed using only one fecal sample [16] . The same result was observed by another group in a recent study assessing the marginal error of Kato-Katz examinations for diagnosis and cure evaluation of S . mansoni infection in areas of low endemicity [17] . Conventional PCR requires several steps after DNA amplification , including electrophoresis or blotting and hybridization , which are limited in the number of samples that can be conveniently analyzed . The PCR-enzyme-linked immunosorbent assay ( PCR-ELISA ) consists of an alternative process for large-scale screening that allows for semi-quantitative analysis . This technique combines an immunological method to quantify the PCR product directly after immobilization of biotinylated DNA on a microplate [18] , [19] . The advantage of PCR-ELISA as compared to PCR-electrophoresis is that it makes use of standard equipment widely used for the processing of ELISAs , and the reagents used are easy to obtain commercially . Therefore , PCR-ELISA allows for the use of PCR-based DNA diagnosis for routine purposes in laboratories in less developed countries with fewer resources . The aim of this work was to design a Schistosoma PCR-ELISA system as an improvement over the PCR assay previously developed by our group to detect S . mansoni DNA in human fecal samples [14] . The performance of the new assay was evaluated with fecal specimens from a Brazilian endemic area for S . mansoni infection and compared with the parasitological Kato-Katz technique for detection and estimation of the intensity of infection .
For use throughout the development of the Schistosoma PCR-ELISA system and for the estimation of its lower detection limit , S . mansoni eggs were obtained from the livers of Swiss albino mice 60 days after infection with 150 cercariae and stored at −20°C in 1 . 7% saline until use [20] . The animals were handled according to local and federal regulations , and the research protocol was approved by the Fiocruz Committee on Animal Research ( License L-0118/09 ) . The number of eggs in saline suspension was quantified using a Neubauer chamber , and a solution containing approximately 2 , 000 eggs was used for DNA extraction . A negative fecal sample ( verified to be free of S . mansoni eggs by the Kato-Katz technique and negative for the Schistosoma DNA detection by the PCR-ELISA system ) was spiked with the egg-saline solution . The egg count was assessed by the Kato-Katz method , and approximately 500 mg of screened feces were used for DNA extraction . Samples of genomic DNA from the human parasites S . haematobium , S . intercalatum and S . japonicum , and also from S . bovis , S . magrebowiei and S . rhodaini were provided by the Laboratório de Parasitologia Celular e Molecular ( Centro de Pesquisas René Rachou , Fiocruz , Brazil ) and used to assess the genus specificity of the Schistosoma PCR-ELISA system . Two hundred and six people from Pedra Preta , Minas Gerais , Brazil , an endemic area for S . mansoni infection , participated in this study . The group was composed of 69 children ( female/male: 33/36; age range of 1–17 years ) and 137 adults ( female/male: 66/71; age range of 18–86 years ) . Thirty-six healthy members of the laboratory staff participated as negative controls throughout the development of the assay . Written informed consent was obtained from all adult participants and from parents or legal guardians of minors . This study was approved by the Ethics Committee of the Centro de Pesquisas René Rachou , Fiocruz , Brazil ( No . 14/2008 ) . For PCR-ELISA , fecal samples were collected and stored at −70°C until DNA extraction . All samples were evaluated for the presence of S . mansoni eggs by the Kato-Katz method . Twelve glass slides ( 41 . 7 mg/smear ) of a single fecal sample were examined , resulting in a total sample weight of about 500 mg . Egg counts were expressed in eggs per gram of feces ( epg ) , using the arithmetic mean of eggs counts obtained from the 12 slides multiplied by 24 [7] . The intensity of infection was calculated as the geometric mean of the individual egg counts . Individuals with positive fecal examination results were treated with a single oral dose of praziquantel ( 50 or 60 mg/kg for adults and children , respectively ) , according to the recommendation of the Brazilian Ministry of Health . Total DNA from 500 mg of each fecal sample and the DNA from a saline solution containing approximately 2 , 000 S . mansoni eggs were extracted using the QIAamp DNA Stool Mini Kit ( Qiagen GmbH , Hilden , Germany ) , according to the manufacturer's recommendations and following the protocols “DNA Isolation from Large Amounts of Stool” and “Isolation of DNA from Stool for Pathogen Detection” . The heating step was performed at 95°C for 20 min to guarantee egg rupture . The concentration and purity of the DNA were determined spectrophotometrically by readings of A260 and A280 ( Eppendorf , Hamburg , Germany ) . The Schistosoma PCR-ELISA system consisted of a biotin 5′-labeled forward primer ( 5′-GATCTGAATCCGACCAACCG-3′ ) , an unlabeled reverse primer ( 5′-ATATTAACGCCCACGCTCTC-3′ ) and the fluorescein 5′-labeled probe ( 5′-TGGTTTCGGAGATACAACGA-3′ ) . The primers used were previously described [14] and designed to amplify a 121-bp tandem repeat DNA sequence ( GenBank accession number M61098 ) found in the genome of S . mansoni [11] . To control for variation in the efficiency of DNA extraction and PCR-amplification , all clinical samples were evaluated with the human beta actin PCR-ELISA system . The primers used were Aco1 ( 5′-ACCTCATGAAGATCCTCACC-3′ ) and Aco2 ( 5′-CCATCTCTTGCTCGAAGTCC-3′ ) , which were previously described to target the fourth exon of the human beta actin gene ( ACTB ) [21] . In this assay , the sense primer ( Aco1 ) was biotinylated at the 5′ end , and a fluorescein 5′-labeled probe ( 5′-TCTCCTTAATGCACGCACG-3′ ) was designed together with the Schistosoma probe described above using the program Primer3-web 0 . 4 . 0 [22] and submitted to homology searches on the National Center for Biotechnology Information website with nucleotide BLAST program using database Nucleotide collection ( nr/nt ) and Megablast option . Amplification primers , biotinylated primers and probes were purchased from Integrated DNA Technologies , Inc . ( Coralville , Iowa , USA ) . For amplification , fecal DNA samples were diluted fivefold , and 2 µl were used as the template . The same volume was used to amplify S . mansoni egg-derived DNA , artificial S . mansoni egg-spiked fecal DNA and DNA from other Schistosoma species . PCR was carried out in a final volume of 20 µl containing 2 µl of GeneAmp 10X PCR Gold Buffer ( 150 mM Tris-HCl , pH 8 . 0 , 500 mM KCl ) , 2 . 0 U of Amplitaq Gold ( Applied Biosystems , Foster City , CA , USA ) , 0 . 1 µg/µl of BSA ( Sigma , St . Louis , MO , USA ) , 0 . 5 µM of each primer , 1 . 5 mM MgCl2 and 200 µM of each deoxynucleoside triphosphate ( Promega , Madison , WI , USA ) . The cycling programs , preceded by 12 min at 95°C to activate the HotStart Taq polymerase , consisted of 15 cycles of 95°C for 1 min , 63°C for 1 min and 72°C for 30 s; 12 cycles of 80°C for 1 min , 63°C for 1 min and 72°C for 30 s and 7 cycles of 80°C for 1 min , 65°C for 1 min and 72°C for 30 s , followed by a final elongation step at 72°C for 7 min . Positive controls based on DNA from S . mansoni eggs were included in all tests . Negative controls containing all of the elements of the reaction mixture except DNA were also included in each PCR assay as surveillance for contamination . A 120-bp segment of the ACTB gene was amplified in a separate tube containing fecal DNA , following the amplification protocol described above , except that the MgCl2 concentration used was 2 mM . The cycling program , preceded by 12 min at 95°C , consisted of 35 cycles of 95°C for 20 s , 60°C for 30 s and 72°C for 1 min , followed by a final elongation step at 72°C for 7 min . The chance of PCR contamination was minimized by physical separation of the starting materials and the amplified products in different rooms; the rooms contained laminar flux chambers with UV light , and sterile , disposable laboratory supplies were used . Data were processed with SPSS statistical software package 13 . 0 for Windows ( SPSS Inc . , Chicago , IL , USA ) and GraphPad Prism 3 . 0 . 3 software ( San Diego , CA , USA ) . All quantitative variables were individually assessed with the one-sample Kolmogorov-Smirnov test for normality . Absorbance readings and arithmetical means of the number of eggs per gram of feces and template DNA concentration ( both transformed into log scale ) were analyzed by Pearson's parametric correlation coefficient or Spearman's nonparametric correlation coefficient . In order to determine the variability of the assays , intra-assay ( repeatability ) and inter-assay ( reproducibility ) precision levels were measured by comparing the means ± the S . D . and reported as coefficients of variation ( CVs , S . D . /mean X 100% ) . Sensitivity , specificity and 95% confidence intervals ( CIs ) were calculated using the OpenEpi Version 2 . 3 program [24] . Agreement beyond chance was assessed using the Kappa index and interpreted according to Landis and Koch [25]: 1 . 00-0 . 81 is excellent , 0 . 80-0 . 61 is good , 0 . 60-0 . 41 is moderate , 0 . 40-0 . 21 is weak and 0 . 20-0 . 0 is negligible . The X2 test was employed for the comparison of proportions . The level of significance was set at P<0 . 05 .
The system's analytical sensitivity was evaluated using eight 10-fold serial dilutions ranging from 1 . 3 ng to 130 ag of genomic DNA extracted from a saline solution containing S . mansoni eggs . The limit of detection was 1 . 3 fg of genomic DNA , equal to that obtained by 6% polyacrylamide gel electrophoresis and silver staining , since the absorbance value for the sample with 130 ag of DNA was equivalent to the PCR negative control ( 0 . 093 and 0 . 064 , respectively ) ( Figure 1A ) . The correlation between the numeric results ( optical density readings ) of the PCR-ELISA and the log number of the template DNA was significant , with a Pearson's correlation coefficient of 0 . 986 ( P<0 . 0001 ) ( Figure 1B ) . The analytical sensitivity of the assay was also evaluated with 10-fold dilutions of DNA extracted from a negative fecal sample spiked with an S . mansoni egg-saline solution determined by the Kato-Katz method to contain 1 , 534 epg . The Schistosoma PCR-ELISA system was consistently able to detect a sample estimated to contain 0 . 1534 epg . The genus specificity of PCR-ELISA was assessed with purified DNA from S . mansoni , S . haematobium , S . bovis , S . intercalatum , S . japonicum , S . magrebowiei and S . rhodaini adult worms . Results ( Figure 2 ) showed a ladder of PCR products due the amplification of the Schistosoma tandem-repeated unit , with the main DNA band of 110 bp present in all samples , as expected , and absorbance readings comparable to the positive control ( S . mansoni DNA ) . To analyze the repeatability of the Schistosoma PCR-ELISA system , three positive and three negative fecal samples from patients were assessed in four replicates in a single run . The intra-assay CVs for absorbance values for the negative samples were: 5 . 6% , 7 . 4% and 8 . 3%; the CVs for the positive samples were 1 . 9% , 3 . 6% and 4 . 2% . In addition , to measure the reproducibility , four replicates of the same samples were assessed on different days in four different assays . The inter-assay variations of absorbance values for the positive DNA samples were 3 . 8% , 7 . 7% and 15 . 9%; the variations for the negative DNA samples were 14 . 2% , 15 . 4% and 18 . 7% . Fecal samples from 206 patients from an area in Brazil endemic for S . mansoni were analyzed by the Schistosoma PCR-ELISA system and also by the parasitological Kato-Katz technique . Comparison of the results obtained by the Schistosoma PCR-ELISA system and the Kato-Katz technique , based on examination of twelve slides ( a total of 500 mg feces ) , is shown in Table 1 . The geometric mean of the number of eggs per gram of feces estimated by the Kato-Katz technique for the positive samples was 18 , which indicates a low intensity of infection [26] . The prevalence observed using the PCR-ELISA system ( 30% ) was higher than that determined with the examination of twelve slides by the Kato-Katz technique ( 18% ) ( X2 = 8 . 81 , P<0 . 003 ) . The Kappa index of 0 . 663 indicates good agreement between the two methods . Analysis of discordant results showed that 25 samples were positive only by the Schistosoma PCR-ELISA system , and one sample was positive only by the Kato-Katz technique . This patient had very low egg output ( 8 epg ) . Table 1 also shows a comparison of the results obtained by the Schistosoma PCR-ELISA system and the Kato-Katz technique , based on examination of two slides ( a total of 83 . 4 mg feces ) , as is routinely done in epidemiological surveys . The Kappa index of 0 . 491 indicates a moderate agreement between the two methods . Analysis of discordant results showed that 36 samples were positive only by the Schistosoma PCR-ELISA system . Diagnostic parameters were calculated by two different approaches: 1 ) taking two-slide Kato-Katz examination as the reference method for comparison or 2 ) considering twelve-slide Kato-Katz examination as the reference method . The sensitivity values of the Schistosoma PCR-ELISA system were high , regardless of the reference considered: 96 . 3% ( 95% CI , 81 . 7–99 . 3 ) for approach 1 and 97 . 4% ( 95% CI 86 . 5–99 . 5 ) for approach 2 . Specificity values changed significantly depending on the reference used , being 79 . 9% ( 95% CI 73 . 4–85 . 1 ) for approach 1 and 85 . 1% ( 95% CI 78 . 7–89 . 7 ) for approach 2 . All 206 fecal samples analyzed by the ACTB PCR-ELISA system showed a positive result , ensuring that negative results correspond to true negative samples for the Schistosoma PCR-ELISA system rather than to a problem with sample degradation or PCR inhibition . Also , the remaining sixteen fecal samples of non-infected persons were evaluated to be positive for the ACTB PCR-ELISA system and negative for the Schistosoma PCR-ELISA system . The ability of the Schistosoma PCR-ELISA system to estimate the parasitic load was assessed , and a Spearman's correlation coefficient of 0 . 616 ( P<0 . 0001 ) was found when comparing values of absorbance readings and epg ( transformed into log[epg+1] ) to those determined by the Kato-Katz technique with 12 examined slides ( Figure 3A ) . When the correspondence between the methods was evaluated , considering only positive and negative samples for both , the correlation coefficient observed was 0 . 700 ( P<0 . 0001 ) ( Figure 3B ) . Results obtained with the Schistosoma PCR-ELISA system based on two levels of intensity of infection ( 1–100 epg and >100 epg ) , according to the Kato-Katz stool examination method , are shown in Table 2 . The high sensitivity of the Schistosoma PCR-ELISA system was evidenced by the high number of positive samples in both groups . The assay also revealed the potential to be considered semi-quantitative , as the mean absorbance readings corresponded to the intensity of infection .
The control of schistosomiasis-related morbidity has become feasible due to the development of single-dose oral drugs such as oxamniquine and praziquantel , which are given to heavily infected patients ( high worm burden ) that were easily detected by field-applicable parasitological methods . With the transition to lower morbidity , there is need for more sensitive diagnostic methods . In an attempt to surpass this diagnostic limitation , a PCR-ELISA system was developed and evaluated as a new molecular assay for the diagnosis of Schistosoma infection . This system showed sound features , such as a low analytical limit of detection , genus specificity ( and absence of cross-reactivity with other related parasites ) , good precision , high sensitivity , good specificity and also the potential for semi-quantitative analysis of parasitic load . Evaluation of the analytical sensitivity of the Schistosoma PCR-ELISA system showed that it could accurately detect 1 . 3 fg of S . mansoni genomic DNA , which is equivalent to less than the DNA found in a single cell of this multicellular parasite , since its genome contains around 580 fg [27] . The high sensitivity of the assay is attainable due to the high copy number of the target sequence , which comprises approximately 12% of the S . mansoni genome [11] . In another approach , the analytical sensitivity was evaluated as the capability to detect DNA samples according to the egg count . The result obtained ( 0 . 1534 epg ) corresponds to fractions of an egg . The Schistosoma PCR-ELISA system was developed using primers designed by Pontes et al . [14] , who demonstrated specificity by the absence of amplification when DNA from four related helminths ( Ascaris lumbricoides , Ancylostoma duodenale , Taenia solium and Trichiuris trichiura ) was used as templates . Further assay also showed no crossreactivity with Fasciola hepatica ( data not shown ) . Additionally , the potential use of the previously described primers and probe and the current PCR protocol to detect genomic DNA from other Schistosoma species was assessed by the amplification of DNA extracted from worms of seven Schistosoma species ( S . mansoni , S . haematobium , S . bovis , S . intercalatum , S . japonicum , S . magrebowiei and S . rhodaini ) . The results obtained confirmed what Hamburger et al . [11] partially addressed using a different set of primers that targets the internal part of the same DNA sequence: namely , this 121-bp tandem repeat is genus specific . The precision of the Schistosoma PCR-ELISA system can be considered good , as the intra-assay variation was lower than 9% , and the inter-assay variation was around 3% to 19% . These percentages are similar to , or lower than , those calculated for other PCR-ELISA assay based on similar principles [28] . The choice of a reference test is crucial for the evaluation of any new testing method . The Kato-Katz technique is considered the test of choice to diagnose schistosomiasis in fecal samples and was chosen as a reference test for comparison with the Schistosoma PCR-ELISA system . Initially , for a traditional approach ( approach 1 in this study ) , the comparison was made using two Kato-Katz slides , corresponding to 83 . 4 mg feces . The PCR-ELISA system was able to detect S . mansoni DNA in 36 fecal samples from patients with negative results by the Kato-Katz technique ( two slides examined ) , demonstrating high sensitivity ( 96 . 3% ) , though lower specificity ( 79 . 9% ) . This specificity is likely to be artificial and incorrect , resulting from the less sensitive and inadequate reference method . In another approach ( approach 2 ) , twelve Kato-Katz slides were evaluated , corresponding to 500 mg of feces , the same quantity used to extract total DNA for analysis with the Schistosoma PCR-ELISA system . Again , the PCR-ELISA system detected more cases of infection with S . mansoni . Thus the sensitivity was high ( 97 . 4% ) , and the specificity was more satisfactory ( 85 . 1% ) . It is well documented that the Kato-Katz method lacks sensitivity if only a single fecal sample is examined , particularly in areas with high proportions of light-intensity infections . A small number of eggs unequally excreted over several days or patchily distributed may not be detected in the small amount of feces examined , negatively impacting on the method's sensitivity . In order to overcome these shortcomings , examination of an increased number of fecal samples as well as in the number of slides per specimen is required . There exists a general consensus that two Kato-Katz slides for each of three fecal samples yield enough sensitivity to obtain reliable data [29]-[32] . As this present study is part of a wider study in the endemic area of Pedra Petra , Minas Gerais , Brazil , three additional samples were collected on different days and analyzed by two Kato-Katz slides each ( data not shown ) . From samples with discordant results between Kato-Katz analysis ( twelve slides examined ) and the Schistosoma PCR-ELISA , 12 out of 25 ( 48% ) were positive in subsequent Kato-Katz examinations of additional samples . Therefore , a better explanation for these discordant results is that the Schistosoma PCR-ELISA system is more sensitive than the Kato-Katz technique considering the same amount of stool examined . That is , these cases correspond to S . mansoni-infected samples and not false-positive results . Although carryover contamination is often difficult to exclude as a cause of false-positive results in PCR-based diagnostic assays , the strict measures taken to avoid it throughout the entire handling and processing steps of the assay make the chances low . Following both approaches , one sample was positive for the Kato-Katz technique and negative for the Schistosoma PCR-ELISA system . Since the parasitological method has an assumed specificity of 100% , this result can be most reasonably attributed to misdiagnosis by the PCR-ELISA . Inhibition of the amplification reaction by fecal compounds and/or DNA degradation during transportation of the sample from the field were considered the most likely technical causes . However , they were subsequently ruled out , as a positive result was obtained with the control ACTB PCR-ELISA system . Therefore , a better explanation seems to be the uneven distribution of a small quantity of eggs in the feces [33] , since the sample had a very low parasite load ( 8 epg ) . ELISA methods for the detection of PCR products provide an alternative to gel-based detection . Among other features , ELISAs prevent possible subjective interpretations of PCR results due to “nonspecific products” or “bands of unknown origin . ” The main advantage of the procedure is the ability to process many samples in parallel ( e . g . , the 96-well microplate format ) , using instrumentation developed for processing ELISA for antibody detection . In addition , the procedure takes less than 2 . 5 hours to be performed ( with the PCR Plate Detection Kit; Sigma ) . The detection format of the commercial kit used here combines methodological features of PCR-ELISA methods described by Landgraf et al . [18] and Luneberg et al . [19] . Biotin-streptavidin binding was used for immobilization of amplification products on microtiter plates , taking advantage of the biotin moiety conjugated to one PCR primer . Subsequent hybridization was carried out with a fluorescein-labeled oligonucleotide probe , with the advantage of confirming the specificity of the test and allowing for detection by an anti-fluorescein-HRP conjugate . Regardless of the fact that real-time PCR provides the best means to quantify PCR products , PCR-ELISA also allows for quantification , although it is usually considered a semi-quantitative technique , as it analyzes end-point amplification products . At the end-point , net synthesis is significantly reduced , inhibitory effects have accumulated and differences in initial starting template concentrations are masked [34] . These problems may be overcome by stopping the reaction in the exponential phase ( as was done in the present study ) or by requiring competitive amplification with very similar template DNA and primer pairs [18] . One classification used for epidemiological estimates of the intensity of infection of schistosomiasis on a community level considers the following groups: i ) light , ≤100 eggs/gram of feces; ii ) moderate , 101 to 400 eggs/gram of feces and iii ) heavy , >400 eggs/gram of feces [26] . Even though the PCR-ELISA system developed in this study showed a satisfactory correlation coefficient between absorbance readings and egg counts per gram of feces , it seemed to be adequate to classify the samples according to the same parameters . However , as the performance of this new assay was evaluated in a low endemic setting , the number of samples with high egg counts was small , limiting statistical analysis . Thus , a statistically significant difference between mean absorbance values was observed only between the positive and negative groups ( P<0 . 0001 for all ) . Nevertheless , an increase in the mean absorbance readings with increasing intensity of infection could be observed . Additional studies in areas of higher prevalence , resulting in a statistically significant number of patients with high egg counts per gram of feces , need to be carried out in order to validate the Schistosoma PCR-ELISA system's semi-quantitative potential . The choice of a molecular method for diagnosing Schistosoma infection in endemic settings will depend on different factors , chiefly the available infrastructure and the cost-effectiveness . A control program may decide between performing multiple sample collections for the Kato-Katz examination or carrying out a single PCR-ELISA test , being the former less costly and the last more accurate . Considering only reagents , the cost of the PCR-ELISA system is roughly US$10 per fecal sample . In conclusion , the Schistosoma PCR–ELISA system constitutes a precise tool for the diagnosis of Schistosoma infection , which may be particularly useful in low-prevalence settings and probably for post-treatment situations . | Schistosomiasis is a neglected disease caused by worms of the genus Schistosoma . The transmission cycle requires contamination of bodies of water by parasite eggs present in excreta , specific snails as intermediate hosts and human contact with water . Fortunately , relatively safe and easily administrable drugs are available and , as the outcome of repeated treatment , a reduction of severe clinical forms and a decrease in the number of infected persons has been reported in endemic areas . The routine method for diagnosis is the microscopic examination but it fails when there are few eggs in the feces , as usually occurs in treated but noncured persons or in areas with low levels of transmission . This study reports the development of the PCR-ELISA system for the detection of Schistosoma DNA in human feces as an alternative approach to diagnose light infections . The system permits the enzymatic amplification of a specific region of the DNA from minute amounts of parasite material . Using the proposed PCR-ELISA approach for the diagnosis of a population in an endemic area in Brazil , 30% were found to be infected , as compared with the 18% found by microscopic fecal examination . Although the technique requires a complex laboratory infrastructure and specific funding it may be used by control programs targeting the elimination of schistosomiasis . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology"
] | 2010 | Development and Evaluation of a Sensitive PCR-ELISA System for Detection of Schistosoma Infection in Feces |
The genus Bartonella comprises facultative intracellular bacteria adapted to mammals , including previously recognized and emerging human pathogens . We report the 2 , 341 , 328 bp genome sequence of Bartonella grahamii , one of the most prevalent Bartonella species in wild rodents . Comparative genomics revealed that rodent-associated Bartonella species have higher copy numbers of genes for putative host-adaptability factors than the related human-specific pathogens . Many of these gene clusters are located in a highly dynamic region of 461 kb . Using hybridization to a microarray designed for the B . grahamii genome , we observed a massive , putatively phage-derived run-off replication of this region . We also identified a novel gene transfer agent , which packages the bacterial genome , with an over-representation of the amplified DNA , in 14 kb pieces . This is the first observation associating the products of run-off replication with a gene transfer agent . Because of the high concentration of gene clusters for host-adaptation proteins in the amplified region , and since the genes encoding the gene transfer agent and the phage origin are well conserved in Bartonella , we hypothesize that these systems are driven by selection . We propose that the coupling of run-off replication with gene transfer agents promotes diversification and rapid spread of host-adaptability factors , facilitating host shifts in Bartonella .
Horizontal gene transfer contributes to phenomena such as pathogen emergence and antibiotics resistance , with major implications for human health . However , little is known about the mechanisms and ecological factors that influence the intensity of horizontal gene transfer in bacteria adapted to wild host populations . With about 170 cases of emerging infectious diseases during the last two decades , more than half of which are caused by bacterial infections [1] , a better understanding of the mechanisms of transfer in natural isolates and how the spread of the mobile gene pool depends on environmental variables such as population size , number of infected hosts and transmission dynamics is needed . Bartonella is a particularly good model system for studies of the mechanisms whereby genes are transferred in wild host populations and the selective forces leading to fixation of the transferred genes in the population . The genus consists of circa 20 described vector-borne species that infect erythrocytes and endothelial cells of mammals [2] , including rodents that are major carriers of infectious disease agents . Four genomes have been sequenced to date , from Bartonella quintana , the agent of trench fever [3] , Bartonella bacilliformis , the agent of Carrion's disease ( unpublished ) , the cat-adapted Bartonella henselae which causes cat-scratch disease during incidental infection in human [3] and the rat-associated Bartonella tribocorum [4] . With single , circular chromosomes in the 1 . 4–2 . 6 Mb range , Bartonella species have the most highly reduced genomes in the bacterial order Rhizobiales of the alpha-proteobacteria . A recent genomic survey identified many putative host-adaptability genes essential for bloodstream infection of B . tribocorum in rats [4] . The experimentally best characterized such genes in B . quintana , B . henselae and B . tribocorum are those coding for the type IV ( VirB , Trw ) and type V ( adhesins , autotransporters ) secretion systems . The VirB system translocates effector proteins into the cytoplasm of endothelial cells , thereby mediating anti-apoptotic activity and angiogenic reprogramming [5]–[7] , whereas the Trw system is required for invasion of erythrocytes [8] and displays diversifying selection on the pilus proteins [9] . The Bartonella adhesin ( BadA ) is involved in binding to endothelial cells , autoaggregation , and mediates a proangiogenic response in B . henselae [10] , [11] . The homologous proteins in B . quintana have essentially the same function , and are required for bloodstream infection , but have been termed Vomps ( variably expressed outer membrane proteins ) since they have been shown to undergo sequence and expression variation in a monkey model system [12]–[14] . Less is known about the autotransporters , however they are known to be upregulated during infection of endothelial cells [15] and may be involved in adhesion to host cells [16] . Several of the genes for secretion systems are located in a dynamic region of the genome , up to several hundred kb in size , which is thought to have originated by the integration of an auxiliary replicon [3] . DNA from this region was highly amplified in B . henselae upon prolonged growth [17] . Microarray hybridizations suggested bi-directional replication starting from a region encoding a few phage genes [17] . A similar phenomenon has also been observed in lambdoid phages of Salmonella [18] , and has been termed escape replication or run-off replication . It has been suggested to arise from defects in prophage excision [18] , [19] . Despite the replication of large amounts of the bacterial chromosome , the prophage that induced escape replication in Salmonella produced phage particles that only contained phage DNA [18] . Bacteriophage particles have been identified in several species of Bartonella [20]–[24] and growth experiments have suggested phage induction and lysis of B . henselae cells as they enter stationary phase [23] . While the morphology of the bacteriophage particles differed slightly – phages from B . bacilliformis and Bartonella vinsonii ssp . berkhoffi were tailed , whereas those from B . henselae and B . quintana lacked tails – they all contained 14 kb linear , double-stranded DNA , packaged in a round to icosahedral head of 40–80 nm in size [21] , [22] , [24] . No one has been able to fully characterize the 14 kb band , however it has been suggested to consist of random or heterogeneous bacterial DNA [21] , [22] , [24] . To learn more about the mechanisms and selective forces driving run-off replication , we determined the genome sequence of Bartonella grahamii strain as4aup , and studied the DNA content of bacteriophage particles from this species . B . grahamii has been isolated from several species of mice and voles and is likely to be one of the most prevalent Bartonella species in wild rodents [25]–[28] . It is transmitted by the rodent flea [29] and has been involved in two reported cases of human disease [30] , [31] . The sequenced strain was isolated from a wood mouse ( Apodemus sylvaticus ) in central Sweden [25] . Through its broad host range , B . grahamii has access to a large gene pool . Since our sequenced isolate has undergone minimal cultivation in the laboratory , we expect it to represent a Bartonella genome in nature , with intact accessory DNA . We demonstrate that the highly dynamic region of the chromosome , that contains many gene clusters for secretion systems , is extensively amplified and packaged into bacteriophage particles . This is the first report that associates run-off replication with bacteriophage particles . We propose that the combination of these two systems promotes diversification and rapid spread of selectively favored host-adaptability genes within and among Bartonella populations , facilitating host shifts .
In this study , we have resolved the long-standing debate about the content of the Bartonella bacteriophage particles . Altogether , our data suggests the presence of two distinct bacteriophages in B . grahamii , one of which is encoded by prophage I and packages its own DNA . The other bacteriophage , which is encoded by genes in phage cluster II , contains chromosomal DNA with an over-representation of DNA from the high plasticity zone . Bacteriophage particles that package random bacterial DNA rather than their own , with no lysis or other obvious negative consequences for the host cell are called gene transfer agents ( GTAs ) . They are usually encoded by a segment of circa 15 kb that contains phage head and tail genes and they transfer DNA in 4–14 kb pieces [36]–[40] . These characteristics agree well with our observations , suggesting that the bacteriophage packaging the 14 kb band is a GTA . To our knowledge , this is the first demonstration of a GTA that packages chromosomal DNA that has first been partially amplified by run-off replication . The identification of a GTA encoded by phage cluster II explains the previously puzzling observation of bacteriophage particles containing 14 kb DNA in many Bartonella species , including B . quintana and B . bacilliformis , that lack prophage I . A gene transfer agent , called RcGTA , which packages the genome in 4 . 5 kb pieces has previously been identified in the alpha-proteobacterial species Rhodobacter capsulatus [36] . Homologous gene clusters are widespread in the alpha-proteobacteria and the congruence in phylogenetic relationships inferred from rRNA and RcGTA genes suggests vertical transmission from a single GTA-containing ancestor [41] . However , no such genes could be identified in any of the Bartonella genomes , although most other members of the Rhizobiales contain partial and/or rearranged RcGTA-like gene clusters . The genes in phage cluster II also show no similarity to other known GTAs , suggesting that the Bartonella GTA is of a novel kind . Despite the absence of sequence similarity between these two GTAs in the alpha-proteobacteria , their overall organization is similar , including genes coding for terminase , portal protein , and putative capsid and tail proteins . We identified gene clusters homologous to the Bartonella GTA in three recently sequenced genomes from the Rhizobiales species Azorhizobium caulinodans , Methylobacterium radiotolerans and Rhodopseudomonas palustris ( Table S2 ) , all of which also contain the RcGTA-like gene cluster . The scattered occurrence of the Bartonella GTA gene clusters in the alpha-proteobacteria indicates horizontal transmission . Could run-off replication be initiated from a plasmid replication initiation site inherited from a once self-replicating megaplasmid ? Such a phenomenon is believed to have occurred in the archaea , where multiple replication initiation sites have been demonstrated [42] , [43] , and suggested to have arisen from the capture of an extra-chromosomal plasmid-derived element [44] . Arguing against a plasmid-derived origin in Bartonella is that in the region close to the peak of the amplification we observed no sequence similarity to the repABC genes , which drives the replication and contains the ori in the megaplasmids of the Rhizobiales [45] . These genes are also not present anywhere else in the amplified region , providing no evidence for a megaplasmid-derived origin of replication . It is more likely that replication starts from an origin derived from an inactivated prophage , as does escape replication in Salmonella . In this species , run-off replication can be induced in the laboratory by inactivating the integrase and/or excisionase genes [18] , showing the ease with which a prophage can lose control of its own replication machinery . Capture of bacteriophage replication machineries has also been observed in mitochondria , where the nuclear genes for DNA polymerase and DNA primase-helicase ( Twinkle ) are derived from T7 bacteriophages [46] . It is interesting to note that the helicase that is located at the peak of run-off replication in Bartonella is homologous to Twinkle , supporting the hypothesis that the extra origin of replication is phage-derived . Homologous genes are present also in several other bacterial genomes , raising the possibility that capture of bacteriophage systems for bacterial replication may occur more frequently than recognized until now . However , no genes for the T7-like DNA or RNA polymerases could be found in the Bartonella genomes , suggesting that replication and transcription of the chromosomal high plasticity zone is not simply driven by a T7-like bacteriophage . An important question is whether run-off replication and GTA-production are two independent phenomena , resulting from an error in the phage replication machinery and a degrading prophage , or whether the systems driving these processes co-evolve under selection and control of the bacterial genome . Several lines of evidence argue in favor of selection , particularly the conservation of phage cluster III , containing the origin of run-off replication , and phage cluster II , encoding the GTA , in all sequenced Bartonella genomes . Normally , prophages degrade rapidly as observed for prophage I , which has been lost independently from B . quintana and individual B . henselae strains [3] , [17] . This is in analogy to the maintenance and vertical inheritance of the RcGTA-like gene cluster among several alpha-proteobacterial species , which has been attributed to selection [41] . What could be the selective advantage of linking run-off replication with a gene transfer agent ? One advantage that we can think of is gene diversification by recombination and rapid spread of new gene variants . Because many genes for secretion systems are located in the amplified region , host adaptation and host switches could select for linkage of the two systems . Indeed , rapid diversification and horizontal transfer of host-adaptability genes within Bartonella populations have been demonstrated for the trw gene cluster [9] . Rodents in particular represent a large and genetically highly divergent host population , and multiple variants of surface proteins could help escape the immune system and promote interactions with a diverse set of host cell molecules . Under this hypothesis , it may be no coincidence that the rodent-associated Bartonella genomes contain the highest number of genes for secretion systems . If run-off replication and production of GTAs are driven by selection , what are the cellular mechanisms and conditions responsible for regulating these processes ? Our observation that the Bartonella GTAs containing amplified DNA are present already during exponential growth phase suggests that these genes are not induced to escape stressed or dying cells , nor that the GTAs are the cause of lysis and cell death . The relatively higher abundance of the products of run-off replication observed previously at later growth phase in B . henselae [17] is probably due to a gradual accumulation of amplified DNA with time . We can also exclude the possibility that run-off replication is regulated by prophage I , in trans , since B . henselae strain GreekCat-23 , which lacks this prophage , still induces run-off replication [17] . Although the Bartonella GTA may very well be under the control of bacterial genes , the regulatory circuits involved are probably different from the growth-dependent production of RcGTA by histidyl-aspartyl signaling proteins [36] as well as by quorum sensing through long-chain acyl-homoserine lactones [47] . An exciting avenue for future work is to investigate how expression of the Bartonella GTA is regulated and whether run-off replication is regulated by the same control systems . It would also be interesting to determine where and when run-off replication and bacteriophage production occur within the rodent host and its fleas and whether different genome variants are generated during the infectious process . The availability of both the B . grahamii and the mouse genome also provides a technical platform that will enable future studies on the co-evolution of bacterial and host genes in natural mouse populations .
Bartonella grahamii strains as4aup and af165up were isolated from a wood mouse and a yellow-necked mouse , respectively , captured in the vicinity of Uppsala in central Sweden [25] . Bartonella henselae strain GreekCat-23 was collected from a cat in the Thessaloniki area in Greece by Aphrodite Tea . The Bartonella strains were routinely grown on hematin agar plates in a humidified 5% CO2 incubator at 35°C . For growth in liquid culture , plate-grown bacteria were inoculated into Schneider's insect medium ( Sigma ) supplemented with 10% FBS and 5% sucrose as described by Riess et al . [48] , additionally supplemented with 25 mM HEPES to keep the pH near 7 . Numbers of viable bacteria were determined as colony-forming units ( CFU ) by plating 10-fold serial dilutions . Starting from single colonies , B . grahamii strain as4aup was cultured on hematin agar plates for five days . Genomic DNA was randomly sheared by nebulization and 1–3 kb sized fragments were recovered . The extracted fragments were cloned into a modified M13 vector using the ‘double adaptor’ method [49] . After 7–8 hours propagation in E . coli , ssM13 DNA was prepared for direct sequencing using Multi Screen MABCN1250 filter plates from Millipore combined with sodium-perchlorate lysis . Direct sequencing of the inserts was performed using DYEnamic ET Terminator Cycle Sequencing Kit ( Amersham ) on the MegaBACE 1000 DNA Analysis System according to the manufacturer's instructions . Cycle sequencing reaction cleanup was performed with the AutoSeq 96 Dye Terminator Clean-up Kit from Amersham . A total of 31 , 166 shotgun sequences were obtained from the M13 library of B . grahamii . After assembly using the Phred and Phrap software [50] , [51] , physical gaps were closed by short and long-range PCR using primers from contig ends . Long-range PCR products were sheared by nebulization and end repaired with DNA Terminator End Repair Kit from Lucigen and blunt-end cloned into pcSmartHCKan vectors using the Clonesmart cloning kit and E . coli XL2-Blue Ultracompetent Cells from Stratagene . Recombinant colonies were grown in 50 µl 2xYT/kanamycin at 37°C overnight without shaking . 1 µl of the culture was used for amplification with TempliPhi Amplification Kit ( GE-Healthcare ) . Both ends of the inserts were sequenced as described above . All ambiguous sites were manually edited with Consed [52] by re-sequencing of PCR products when necessary . A number of polymorphic sites in the badA region remain unresolved , these may well represent authentic polymorphisms since this gene is known to evolve very rapidly but we cannot exclude the possibility that there should be an additional badA gene . Bacteria were grown on hematin agar plates for 10 days . Prior to DNA extraction , the cells were suspended in TNE buffer ( 10 mM TRIS pH 8 . 0 , 150 mM NaCl and 1 mM EDTA ) and centrifuged; washes were repeated twice . DNA was isolated in agarose plugs , made using 2% SeaPlaque GTG agarose ( Cambrex Bio Science , ME , USA ) in 0 . 5× TBE buffer as described in [53] , with minor modifications . One mm thick slices of DNA-containing plugs were digested with 10 U of NotI , AscI , RsrII , SbfI ( New England Biolabs ) and SgfI ( Promega ) restriction endonucleases separately , after 30 min pre-equilibration in TBE buffer on ice and subsequent 30 min pre-equilibration with an appropriate restriction buffer . Restriction was performed overnight at recommended conditions . The DNA fragments were separated in 0 . 9% PFGE-grade agarose ( SeaKem Gold; Cambrex Bio Science ) in 0 . 5× TBE buffer in the GenNavigator System apparatus ( Amersham Biosciences ) at 14°C and 5 . 6 V/cm , for a total of 65 hours . The total run was separated in six phases , using switch times ramped from 5 to 150 s . The sizes of the fragments were estimated using PFGE l-ladder and Yeast Chromosome PFG marker ( New England Biolabs ) . Thirty-seven DNA fragments serving as probes were amplified , purified and labeled as described earlier [17] . Southern blotting , hybridization and signal detection were performed as described earlier [17] . NotI restriction digest was used to verify the assembly of the B . grahamii genome . DNA was digested and separated as described above , using 1 . 1% low melting point SeaPlaque agarose for casting the gel . The gel was run and bands were excised as suggested by the protocol extraction of high molecular weight DNA from gel [54] . The blocks of gel containing individual bands of interest were excised and the agarose was digested using ß-agarase I ( New England Biolabs ) according to the manufacturer's instructions . From the reaction mix , DNA was extracted with equal volume of phenol: chloroform: isoamyl alcohol ( 25∶24∶1 ) . The aqueous phase was cleaned up with equal volume of chloroform: isoamyl alcohol ( 24∶1 ) , and DNA was precipitated with 2 volumes ice-cold isopropanol at −20°C overnight , washed with 70% ethanol twice and re-dissolved in TE buffer pH 7 . 5 . 1 µg DNA from each of the two smallest NotI bands was hybridized against total genomic DNA to a microarray ( see below ) , to see which part of the genome the band represented . Curation and annotation of the genome were done with the annotation platform GenDB [55] . Protein-coding genes were predicted with GLIMMER [56] and CRITICA [57] . Similarity searches were performed against several databases , including GenBank ( nt and nr ) , UNIPROT [58] , KEGG [59] and COG [60] . Protein domains were identified with InterProScan [61] and the InterPro database [62] . SignalP [63] , helix-turn-helix [64] , and TMHMM [65] were used . RNA genes were identified with tRNA-Scan-SE [66] and ARAGORN [67] . Orthologs between Bartonella species were predicted as reciprocal best Blastp [68] hits , excluding genes with top hits in the same species . Paralogs were predicted using blastclust ( http://www . ncbi . nlm . nih . gov/Web/Newsltr/Spring04/blastlab . html ) , requiring 50% similarity over 40% of the length . Phage genes were defined based on Blast hits and literature . Genes were defined as orphans if there was no convincing Blast hit in the GenBank nr database . Genes were defined as Bartonella-specific if the only hits were to other Bartonella species . For remaining genes , all orthologs and paralogs along with Blast hits from a selection of 27 completely sequenced alpha-proteobacterial genomes were used for phylogenetic analysis . To get a broad representation of species we further selected top Blastp hits up to a total of 40 genes ( E<1e−10 ) , excluding all other alpha-proteobacterial hits . Protein sequences were aligned with clustalW [69] , and phylogenetic trees were inferred with PHYML [70] , using the JTT model of amino acid substitution , fixed proportion of variable sites , one substitution rate category , estimated gamma distribution parameter and 100 bootstrap replicates . To get a rough estimation of the number of imported and vertically inherited genes , we inspected all trees with less than 18 Blast hits to the selected alpha-proteobacterial genomes manually . Imported genes were defined as those for which there was no indication of vertical inheritance from the ancestor of Bartonella and its close relatives . The genes for Type IV secretion systems were classified as imported despite being common in the alpha-proteobacteria , since previous analyses suggest a recent acquisition of these genes [33] . Gene classifications were transferred from B . grahamii to the other Bartonella species using best Blast hit ( E<0 . 001 ) . Genes in the other species without homologs in B . grahamii were manually classified . Genomic islands were manually located in B . grahamii based on a synteny comparison with the previously published genomes ( B . tribocorum , B . henselae and B . quintana ) . A genomic island was defined as a contiguous segment larger than 10 kb containing orphans , non-orthologous genes , or long non-coding regions . Island boundaries were extended to include adjacent Bartonella-specific genes or imported genes . For simplicity , boundaries of islands were always defined as the boundaries of the adjacent non-island genes . Protein sequences for FHA and autotransporters were aligned with Kalign [71] and phylogenetic trees were inferred with RAxML [72] , substitution model PROTMIXWAG , 100 bootstrap replicates . Since the autotransporter genes are of very variable length , only the conserved beta-domain , identified with HMMPfam in InterProScan [73] , was used for phylogenetic inference . Bacterial cells were either harvested after 5 to 15 days of culture on hematin agar plates , or at selected time points during growth in liquid culture . While samples from liquid culture were collected directly , plate-grown bacteria were either overlaid with 5 ml Brain-Heart Infusion Broth ( Difco ) the day prior to collection or re-suspended in SM buffer . After centrifugation at ca . 4000×g for 15 min at 4°C , supernatants were collected , filtered through 0 . 2 µm filters and enriched using polyethylene glycol as described in [21] . Phage pellets were re-suspended in 30 µl TE buffer pH 8 . 0 or PBS . Extraction of DNA from the bacterial pellets was performed using the AquaPure Genomic DNA Kit ( Bio-Rad ) according to the manufacturer's instructions . Phage DNA from plate cultures was extracted as described in [21] . Phage DNA from liquid culture was extracted in combination with amplification using the REPLI-g Mini kit ( Qiagen ) according to the manufacturer's protocol for DNA amplification from blood or cells . Concentrations were determined by measuring absorbance using a NanoDrop ND-1000 . Size selection of phage DNA fragments from non-amplified preparations was carried out by separation in 0 . 7% ( w/v ) low melting point agarose ( SeaKem Plaque , Cambrex ) gels containing ethidium bromide , in 1× TAE buffer at 4°C at 6 v/cm overnight . The fragments were visualized under near-UV light ( λ = 360 nm ) and excised with a sterile scalpel blade . Two volumes of ß-agarase I buffer ( New England Biolabs ) amended with 100 mM NaCl , 30 µM spermine and 70 µM spermidine [58] were added to each slice , followed by equilibration on ice three times for 30 min . Gel slices were melted at 65°C for 15 min , cooled to 42°C for 10 min and digested with 2 U of ß-agarase I ( New England Biolabs ) for one hour . ß-agarase was deactivated at 95°C for 5 minutes and the DNA was extracted by equal volume of phenol-chloroform-isoamyl alcohol ( 25∶24∶1 ) as described in [54] . DNA from the 14 kb band of B . grahamii af165up was amplified using the REPLI-g Mini kit . The protein components of the bacteriophage particles were separated by SDS-PAGE ( 12% ) after being heat-denaturized at 95°C for 5 minutes . Mass spectrometry was performed for each band . Proteins were identified based on computational searches in the NCBI non-redundant ( nr ) database and a local database consisting of the B . grahamii protein sequences . We designed 4438 60-mer oligonucleotides , representing 1703 protein-coding genes , 110 pseudogenes and 663 spacers with OligoArray2 . 1 [74] . A probe was considered to belong to a gene if at least half the probe was located within the predicted borders of the gene . The probes were spotted in three replicates on Ultra-GAPS-coated slides ( Cornings , Inc . ) . Slides were cross-linked at 250 mJ/cm2 , and prehybridized at 42°C for 45–60 min in 3X SSC , 0 . 1% SDS and 0 . 1 mg/ml bovine serum albumin . Labeling of genomic DNA was performed as in [75] . Hybridizations were performed at 42°C overnight in hybridization solution containing 5X SSC , 0 . 1% SDS , 0 . 1 mg/ml sonicated salmon sperm DNA and 20% formamide . Scanning and image analysis were performed with a GenePix 4100A scanner and the GenePix 5 . 1 software ( Axon Instruments , Molecular Devices ) as described previously [17] . Phage and bacterial DNA are referred to as Ch1 and Ch2 respectively , and spots meeting the following criteria were removed from further analysis: Spots flagged as bad or not found during quantification , spots with more than 5% saturated pixels in either channel , Ch1 spot median below 3 times the Ch1 background median , less than 95% of pixels having a Ch1 intensity higher than background intensity plus one standard deviation , less than 90% of pixels having a Ch1 intensity higher than background intensity plus two standard deviations or spots having less than 70 pixels . M values were computed as log2 ( Ch1/Ch2 ) . The genome sequence of B . grahamii has been deposited in GenBank under the accession numbers CP001562 for the chromosome and CP001563 for pBGR3 . The microarray data have been deposited in the ArrayExpress database of the European Bioinformatics Institute under the accession numbers A-MEXP-1576 for the array design and E-MEXP-2149 for the experimental data . | Emerging infectious diseases represent an increasing human health problem with many examples of disease outbreaks caused by transmissions from animals to humans , such as , most recently , the bird flu virus . Genes involved in virulence and antibiotic resistance are often carried by mobile elements like plasmids and viruses , which mediate transfer between cells at an amazing speed . Rodents represent a major carrier of infectious agents , and it is therefore particularly important to study the gene transfer processes in bacteria that use rodents as their natural host reservoir . We have studied the genome of a bacterium that is naturally adapted to mice and identified many more putative host-interaction genes than were observed in previously recognized human pathogens . Furthermore , most of these genes are located in a segment of about 25% of the genome , which was massively amplified and packaged into viral particles . This is the first demonstration of targeted packaging of a portion of the bacterial chromosome into viral particles , and we propose that this is a novel strategy for increased exchange of genes involved in the infectious process . | [
"Abstract",
"Introduction",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microbiology/microbial",
"evolution",
"and",
"genomics",
"genetics",
"and",
"genomics/comparative",
"genomics",
"evolutionary",
"biology/genomics"
] | 2009 | Run-Off Replication of Host-Adaptability Genes Is Associated with Gene Transfer Agents in the Genome of Mouse-Infecting Bartonella grahamii |
Identifying enhancers regulating gene expression remains an important and challenging task . While recent sequencing-based methods provide epigenomic characteristics that correlate well with enhancer activity , it remains onerous to comprehensively identify all enhancers across development . Here we introduce a computational framework to identify tissue-specific enhancers evolving under purifying selection . First , we incorporate high-confidence binding site predictions with target gene functional enrichment analysis to identify transcription factors ( TFs ) likely functioning in a particular context . We then search the genome for clusters of binding sites for these TFs , overcoming previous constraints associated with biased manual curation of TFs or enhancers . Applying our method to the placenta , we find 33 known and implicate 17 novel TFs in placental function , and discover 2 , 216 putative placenta enhancers . Using luciferase reporter assays , 31/36 ( 86% ) tested candidates drive activity in placental cells . Our predictions agree well with recent epigenomic data in human and mouse , yet over half our loci , including 7/8 ( 87% ) tested regions , are novel . Finally , we establish that our method is generalizable by applying it to 5 additional tissues: heart , pancreas , blood vessel , bone marrow , and liver .
Transcriptional regulation in mammals is a highly orchestrated process directed in part by the binding of sequence-specific transcription factors ( TFs ) to genomic regulatory elements , such as enhancers . Enhancers contain binding sites for sequence-specific TFs that recognize particular DNA motifs . The combined input of the multiple TFs that bind to a single enhancer region results in tissue- and time-point- specific gene activation [1] . Identification of active enhancers , particularly those enhancers that are most relevant to a developmental process , is a challenging task that is the subject of intense investigation . The ENCODE project and Roadmap Epigenomics project have recently provided DNase I hypersensitive sites ( DHSs ) , which can mark enhancers , promoters , silencers , insulators , and locus control regions in many human cell and tissue samples [2] , [3] . Additionally , the mouse ENCODE project has provided ChIP-Seq data for enhancer-associated chromatin marks in multiple mouse tissues and cell types [4] . While highly valuable , these data provide only indirect evidence of cis-regulatory activity . Chromatin must be open for most trans-factors to bind , but not all open chromatin must be active . Other epigenomic marks are highly correlated with characterized cis-regulatory elements , but they are not confined to demarcate only these elements , nor do they mark all of them . Computational analysis can provide valuable complementary information: it can predict the identity of the trans-factors binding to putative cis-regulatory elements , it can highlight enhancers under active purifying selection , and it can be used to provide enhancer predictions in spatio-temporal contexts that have yet to be assayed . Many computational screens have been carried out in an attempt to identify enhancers that are active in a particular tissue [5] , [6] . Previous computational methods often rely heavily on manual curation of TFs that are known to have a role in a particular tissue , or manual curation of lists of known active enhancers [5] . Known enhancers can be used to build a training set that will allow the identification of patterns that are enriched in the training set compared to a background set . Any region across the genome with the same ( binding site ) patterns are putative enhancers in the tissue from which the training set was built [5] . Because these methods rely on manual curation of data sets , they either do not allow discovery of TFs that are important for a process but have not yet been characterized , or are often limited by the enhancer regions they were trained on . Here , we introduce an integrated computational framework to identify enhancers in a specific tissue by searching for clusters of TF binding sites ( TFBS ) with a related function . Our framework first uses a recently published approach to predict high-confidence binding sites across the genome [7] . Then , each TF is associated with biological functions by taking the set of predictions and analyzing them with GREAT , a functional enrichment analysis tool that assigns biological meaning to a set of putative cis-regulatory genomic regions [8] . We use this approach to first identify TFs with functions related to a particular tissue , which solves the constraint of manual literature curation of TFs and allows identification of TFs with previously uncharacterized roles in the tissue . Because transcription factors generally work in concert through discrete enhancer modules [1] , we then search for clusters of binding site predictions for TFs with a related function . These clusters of binding site predictions represent putative enhancers in the tissue of interest . We applied the above method to discover active enhancers in the mammalian placenta , a tissue that is understudied despite its critical role in human development . Placenta development is a complex , step-wise process , where spatio-temporal control of gene expression must be tightly regulated to ensure proper embryonic and fetal growth [9] . In early stages of gestation , trophoblast cells that surround the developing embryo are directly involved in implantation by attaching the blastocyst to the uterine epithelium [9] . As the placenta continues to develop , it contributes to establishing blood flow between mother and fetus , transporting nutrients , and eliminating waste products [9] , [10] . Therefore , distinct genetic programs are activated at various times and locations throughout placenta development . Defects in placenta development have also been associated with human disorders such as preeclampsia , and while many SNPs have been identified in association with preeclampsia [11] , the function of these SNPs remains unknown . Our screen identified 2 , 216 putative placenta enhancers , or TFBS clusters . Of these putative enhancers , 36 were tested using luciferase reporter assays in two placental cell types: mouse trophoblast stem cells ( TSCs ) , and mouse trophoblast giant cells ( TGCs ) differentiated in culture from TSCs . We also tested the candidates in a primary non-placental cell type as a negative control . We found that 31 ( 86% ) of the candidates were able to drive activity in at least one of TSCs and TGCs , the bulk of which had significantly higher activity in trophoblast cells compared to the other cell type . These results show that our method is able to accurately predict evolutionarily conserved placenta enhancers , which likely function in the development of the human placenta . Because our approach is fully integrated with existing gene ontology databases , we demonstrate it can be easily adapted to well-annotated tissue types by running it on 5 additional tissues .
We have previously shown that we can predict TFBS with high accuracy across the genome using an excess conservation metric [7] ( Figure 1A ) . This metric , which improves state of the art TFBS predictions , measures the likelihood for a binding site to be conserved to the observed phylogenetic depth in a particular region of the genome , and favors binding sites that are conserved more strongly than the surrounding sequence [7] . We have also shown that binding site predictions for each TF can be analyzed using a functional enrichment analysis tool , GREAT ( the Genomic Regions Enrichment of Annotations Tool ) [8] , to predict functions for the TFs [7] . GREAT contains terms , or lists of genes that have functional commonalities ( e . g . placenta development ) . Given a particular term , GREAT computes the fraction of the genome covered by the regulatory domains of the genes in the list , and the number of binding site predictions hitting these regulatory domains . These data can be used to calculate a p-value for each term using the binomial test , thereby providing a statistic for the enrichment of TFBS near genes annotated for the particular term . Here we focus the above approach on a particular tissue , the placenta , and search for TFs ( motifs ) most associated with specific GREAT terms related to the placenta , such as “abnormal placenta morphology” , “placenta development” , and “abnormal placenta labyrinth morphology” ( see Materials and Methods ) . We used 917 non-redundant motifs curated from UniPROBE [12] , JASPAR [13] , and TRANSFAC [14] , and used the excess conservation metric to identify high confidence matches to each of the motifs genome-wide , that are conserved in mouse and human . By requiring conservation of predictions between mouse and human , we focus on similarities between the species , which aids the study of human development by use of a mouse model . We then obtained functional enrichments for each of the TFs by analyzing the top 10 , 000 predictions for each motif using GREAT , which allowed identification of the TFs that have the most enrichment for placenta terms ( Figure 1A ) . We next collapsed similar binding motifs ( see Materials and Methods ) , such that each distinct motif was assigned both a q-value ( corrected p-value ) , based on the placenta term that was most enriched , and a most likely TF , based on placenta gene annotations . The 50 TFs with the most significant q-value for a placenta term are shown in Figure 1B . To assess the quality of the 50 TFs predicted to be important for placenta development , we first used an automated method to determine which TFs in the entire ranked list have known roles in placenta development . We predict a TF is involved in placenta development if it has multiple binding sites near genes involved in placenta development , and our prediction can be confirmed when the TF itself is already associated with a placenta term ( Figure 1A ) . To obtain a list of TFs annotated for placenta function , we combined gene lists from two placenta terms in GREAT ( see Materials and Methods ) . We found that a significant number ( p = 0 . 013 ) of TFs in the top 50 appeared in the known placenta gene list ( Supplementary Figure S1 ) . We further assessed the ranked list by manually annotating the top 50 TFs . While gene ontologies can be used to identify many genes associated with a process , they may not identify all genes associated with a process . We classified each TF as either previously known to have a role in placenta function , based on the literature , or predicted to have a role in placenta function , based on our current approach ( Figure 1B , Supplementary Table S1 ) . We determined that 33 ( 66% ) of the top 50 TFs have a known role in the placenta . At the top of our prediction list is Rbpj , for which knockout mice show a number of abnormalities , including defects in placenta development [15] . Other well-known TFs in the top 50 that have knockout mice showing placental defects include Ets2 , Junb , Ascl2 , and Foxo1 [16]–[20] . Interestingly , 17 ( 34% ) of the top 50 TFs do not currently have a well-characterized role in placenta development . We determined the expression levels of the predicted TFs , using published human RNA-Seq data in three placental components: amnion , chorion , and decidua [21] , as well as published mouse RNA-Seq data in TSCs [22] . Fifteen out of the seventeen predicted TFs ( 88% ) are highly expressed in at least one of the four placental cell types ( Supplementary Table S2 ) , which is significant ( p = 5×10−3 ) when compared to the number of TFs highly expressed when TFs are chosen randomly from all TFs with expression values ( 1 , 000 simulations ) , providing further evidence for their role in placenta development . To identify placenta enhancers , we developed an algorithm to discern clusters of binding sites using predictions from the top 50 TFs annotated with a placenta function . The algorithm first uses spatial hierarchical clustering based on distance between predicted TFBS , and then segments the cluster hierarchy based on cluster score ( see Materials and Methods ) . The clustering process groups the initial set of 485 , 038 binding site predictions ( ∼10 , 000 per TF in top 50 ) to 255 , 209 regions ( average length 76 bp , maximum 544 bp ) with 1 or more binding site prediction from the top 50 TFs . The 255 , 209 regions had a low fold enrichment for placenta terms , as determined through GREAT . The fold enrichment reported by GREAT measures the number of regions associated with a term compared to the expected region hits , given the size of the input set and the fraction of the genome covered by the term . Because functional enhancers often harbor binding sites for multiple TFs [1] , we enriched for regions that are likely to function as placenta enhancers by keeping only those regions with 5 or more non-overlapping binding site predictions for one or more of the top 50 TFs , giving 3 , 014 potential placenta enhancers that had an average size of 279 bp . This filtering step increased the fold enrichment for multiple placenta terms well beyond the default GREAT two fold significance threshold ( Figure 2A ) . We further quantified the enrichment obtained by choosing regions with ≥5 TFBS . We show that when randomly choosing 3 , 014 regions from the set of 255 , 209 clusters that contain one or more binding site predictions , the observed q-value ( when regions with ≥5 TFBS are chosen ) of 3 . 99×10−24 for the “placenta development” term is highly significant ( log10 of Z-score = 19 . 08 ) , as the best q-value observed in 1 , 000 simulations was 3 . 85×10−9 ( Figure 2B ) . We also conducted a sensitivity analysis to determine the effect of varying the thresholds we used in this process . We wanted to choose a threshold that would result in a large number of TFBS clusters ( >1 , 000 ) and a high GREAT fold enrichment for a representative GREAT term ( >3 ) . First , we determined the effect on the GREAT fold enrichment for placenta terms when , instead of clustering predictions from the top 50 TFs , we clustered predictions from the top 25 , 75 , or 100 TFs . As shown in Supplementary Figure S2 , using the top 50 TFs provides both the best fold enrichment and best balance between quantity and purity . We next varied the number of non-overlapping TFBS used to identify placenta TFBS clusters from at least 3 non-overlapping placenta TFBS to at least 7 non-overlapping placenta TFBS . Requiring at least 6 non-overlapping placenta TFBS slightly increases the fold enrichment , but provides less than 1 , 000 predictions , while dropping the threshold to at least 4 non-overlapping placenta TFBS lowers the fold enrichment to below 3 ( Supplementary Figure S2 ) . Finally , we tested the effect of including non-placenta TFBS when counting non-overlapping binding sites in a TFBS cluster to accommodate for both general purpose and specific TFBS in the same enhancer . For example , for each threshold of at least 4 non-overlapping TFBS to at least 7 non-overlapping TFBS in a region , we required that ≥3 of the TFBS be from the placenta ( top 50 ) TFs . As expected , we see that reducing the number of TFBS that are required to be from placenta TFs increases the number of TFBS clusters identified . However , it also reduces the GREAT placenta term fold enrichment ( Supplementary Figure S2 ) , indicating the TFBS clusters identified are less likely to be involved in placenta functions . To further enrich for enhancers likely specific to placenta development , we filtered regions that contain a high number of non-placenta TFBS clusters . While these regions may be active enhancers in the placenta , it is difficult to claim that they are specific to the placenta based on TFBS composition alone , as the number of non-placenta TFBS clusters in these regions is high . To this end , we ran the binding site clustering approach using 50 random TFs that were unlikely to have a role in placenta development ( rank below 100 from the ordered list of TFs ) . We carried out this process a total of 1 , 000 times to obtain a set of background clusters with ≥5 non-overlapping TFBS . We then determined the number of times each of the 3 , 014 putative placenta enhancers overlapped a background cluster . Finally , we removed from the putative placenta enhancer set those regions that were identified in at least 5% of the background runs . This process further enriched our set for placenta functions , increasing the range of GREAT fold enrichments for placenta terms to 3 . 17–4 . 56 , and left us with our final set of 2 , 216 placenta TFBS clusters ( Figure 2A , Supplementary Table S3 ) . Because of the initial conservation criteria we used for binding site predictions , each of the placenta TFBS clusters is conserved in human . A roughly balanced 1 . 4% to 8 . 2% of the total predictions we started with for each of the top 50 TFs before clustering ends up in a placenta TFBS cluster; placenta TFBS clusters are generally heterotypic , made up of binding site predictions for multiple different TFs; and all of the top 50 TFs contribute to the different clusters , irrespective of their motif information content , or whether they are already known or not in placental contexts ( Supplementary Figure S3 ) . To functionally test the placenta TFBS clusters , we performed enhancer reporter assays in a mouse placenta cell culture system . This system allows one to maintain trophoblast stem cells ( TSCs ) derived from early mouse development , which are the precursor cells of the differentiated cells of the placenta [23] . TSCs can also be differentiated into trophoblast giant cells ( TGCs ) , which are the placental cell type that invade maternal tissue to help establish maternal-fetal blood flow [23] . This cell culture system therefore allows testing of enhancer elements using luciferase reporter assays in two different placental cell types ( Supplementary Figure S4 ) . Because TSCs often spontaneously differentiate into TGCs , we ensured the purity of our transfected cell populations by using non-overlapping transfection conditions that were optimized for each cell type ( see Materials and Methods ) , and by transfecting the mouse placental lactogen II ( mPL-II ) enhancer as a control for each set of TSC or TGC transfections . PL-II is known to be specific to TGCs , and the enhancer region we tested has been shown to be active in rat trophoblast giant cell-like cells [24] . Indeed , we found consistently high mPL-II enhancer activity in our TGCs and little to no activity in our TSCs ( Supplementary Figure S4 ) . We tested 36 placenta TFBS clusters upstream of a minimal promoter driving luciferase activity in TSCs and TGCs ( Supplementary Table S4 ) . The candidates were selected to cover a range of distances from gene transcription start sites ( TSSs ) , to cover a range of non-overlapping TFBS predictions , and to contain different proportions of binding sites for known to predicted TFs . Luciferase fold activity for all candidates was calculated compared to an empty vector control . Of the 36 candidates , 31 ( 86% ) showed more than 2-fold activity in at least one of TSCs and TGCs , 26 ( 72% ) showed more than 2-fold activity in TSCs , and 28 ( 78% ) showed more than 2-fold activity in TGCs ( Supplementary Figure S5 ) . 19 of the 36 candidates consisted of more binding sites for TFs we predicted to have a role in placenta development rather than previously characterized placenta TFs . Of these 19 candidates , 16 ( 84% ) had activity in at least one of TSCs and TGCs . Additionally , there was no strong correlation between fold activity and proportion of known placenta TFBS in either of the two cells types for the candidates ( TSC r2 = 0 . 19 , TGC r2 = 0 . 17 ) , further implicating the predicted TFs in placenta development . We then tested the specificity of the transfection system in placental cells using 15 negative controls . Because our lab has interest in the neocortex [25] , we chose 6 negative controls that are robust enhancers in neocortical cells ( Supplementary Figure S6 ) . We also chose 9 negative controls within the same range of GC content , length , and level of conservation to the placenta TFBS cluster candidates . For both sets of negative controls , we observed significantly lower activity in TSCs ( unpaired t-test p-value = 7×10−3 ) and TGCs ( unpaired t-test p-value ≤5×10−3 ) compared to the placenta TFBS cluster candidates , demonstrating the robustness of the transfection system ( Figure 3A–B , Supplementary Figure S6 ) . To assess the specificity of the 36 placenta TFBS cluster candidates tested in placental cell types , we also performed luciferase reporter assays on neocortical cells isolated from e14 . 5 mice [25] . Of the 31 candidates that had more than 2-fold activity in TSCs or TGCs , 22 ( 71% ) had significantly higher ( unpaired t-test , p-value <0 . 05 ) activity in TSCs or TGCs compared to neocortical cells , including 12 candidates ( 39% ) that were inactive ( <2-fold activity ) in neocortical cells ( Figure 3C–D and Supplementary Figure S7 ) . To ensure the difference we see is not due to high basal activity of the promoter in placental cell types compared to neocortex , we show that the basal activity of the promoter is low in each cell type and should have no impact on our results ( Supplementary Figure S7 ) . These data demonstrate that our approach is able to identify placenta enhancers with high accuracy , the bulk of which are more active in trophoblast cells compared to neocortical cells . To compare the predicted placenta TFBS clusters more globally to putative enhancers in other tissues , we overlapped them with epigenomic enhancer-associated marks in 18 other mouse tissues and cell types [4] . Each cell type we compared to has , on average , 50 , 166 putative enhancers . We see that 793 ( 36% ) of the placenta TFBS clusters overlap with putative enhancers in zero other tissues . 1 , 108 ( 50% ) of the placenta TFBS clusters overlap with putative enhancers in ≤2 other tissues , and only one placenta TFBS cluster overlaps with putative enhancers in all other tissues ( Supplementary Figure S8 ) . This comparison further suggests that the placenta TFBS clusters are less likely to be active in many other tissues . Our method has led to the identification of previously uncharacterized enhancers in the regulatory domains of target genes that play an important role in placenta development . For example , Hand1 has been shown to be essential for placenta development , as the knockout mice arrest by e7 . 5 and have defective trophoblast giant cell differentiation [26] . Our analysis reveals multiple putative enhancers in the regulatory domain of Hand1 , two of which are over 30 kb upstream of the transcriptional start site and are conserved amongst placental mammals ( Figure 4A ) . Our luciferase reporter assays show that both candidates have significantly higher activity in the placental cell types compared to neocortical cells ( Supplementary Figure S5 ) . Another example is a putative enhancer ∼13 kb upstream of Dll4 , a gene that is involved in the development of the placenta vasculature ( Figure 4B ) [27] , [28] . This candidate was also found to have significantly higher activity in placental cell types than in neocortical cells ( Supplementary Figure S5 ) . In total , 14 of the 36 candidates tested are found in the regulatory domain of genes that are known to be involved in placenta development . Both TSCs and TGCs showed no significant difference in activity for candidates near target genes known to be involved in placenta development compared to those that have unknown functions in placenta ( unpaired t-test , TSC p-value = 0 . 27 , TGC p-value = 0 . 32 ) , suggesting that our method can be used not only to predict TFs but also to predict target genes that were not previously known to have a function in placenta . Because genes important in a particular context can be regulated by multiple enhancers [25] , [29]–[32] , we searched for genes with the most predicted placenta TFBS clusters in their regulatory domains . To carry out this analysis , we determined a q-value ( using GREAT ) for all the genes in the genome based on the likelihood associated with the observed number of placenta TFBS clusters per gene , normalized to the length of the individual gene's regulatory domain . The ten genes with the most significant q-values are shown in Supplementary Table S5 . Of the top ten genes , four of them , Pdgfb , Junb , Epha2 , and Socs3 have previously characterized roles in placenta development . Interestingly , Zbtb7b , a TF we predict to have a role in placenta development based on our approach , is also within the top ten , with five placenta TFBS clusters within its regulatory domain . We next sought to further compare our computational enhancer predictions to published datasets that use biochemical assays to predict enhancers genome-wide . The mouse ENCODE project has recently generated ChIP-Seq data for enhancer-associated chromatin marks in mouse term placenta [4] , and the ENCODE and Roadmap Epigenomics projects have generated DNase-Seq data in human placenta tissue at 85–113 days gestation to assay for open chromatin [2] , [3] . To enrich for enhancer-associated chromatin from the mouse ChIP-Seq data , we combined regions marked by enhancer-associated marks H3K27ac and H3K4me1 that do not contain the promoter-associated mark H3K4me3 . For the human DNase-Seq set , we combined data from six biological replicates ( see Materials and Methods ) . Because we were interested specifically in comparing enhancer-associated regions , we removed regions within 1 kb of gene transcriptional start sites from all three sets before comparison . This brought the number of: placenta TFBS clusters to 1 , 847 ( covering 0 . 02% of the genome ) , mouse histone mark based data to 70 , 951 ( covering 10 . 22% of the genome ) , and human open chromatin data to 80 , 922 ( covering 0 . 38% of the genome ) . We first determined the enrichment for placenta terms using GREAT for all three sets . As expected the placenta TFBS clusters have the highest enrichment for several placenta terms , partly because the process used to define them specifically enriches for a subset of these terms . However , significance for the other two sets was quite low , with most values between 1 . 7–1 . 9 fold enrichment , below GREAT's standard significance cut-off of 2-fold ( Table 1 ) . These low values suggest that it might be more difficult to identify functional placenta enhancers from these large sets . Second , we wanted to determine if any of the placenta TFBS clusters overlapped with the mouse or human enhancer-associated regions . To compare mouse and human data , we first converted human region coordinates to mouse coordinates ( see Materials and Methods ) . We found that 880 ( 48% ) of the placenta TFBS clusters overlapped with the mouse experimental set , and 390 ( 21% ) of the placenta TFBS clusters overlapped with the human experimental set ( Figure 5A–B ) . To determine if the overlaps we observed were significant , we chose 1 , 847 ( the set size of the placenta TFBS clusters that are not within 1 kb of a TSS ) size-matched random genomic regions and checked the overlap with the mouse and human experimental sets . We ran this process a total of 10 , 000 times . The simulation demonstrates that the placenta TFBS clusters have very significant overlap with both the mouse ( p<10−4 , Z-score = 50 . 25 ) and human ( p<10−4 , Z-score = 77 . 71 ) experimental data ( Figure 5C–D ) . The simulation also demonstrates that while the human experimental set has less overlap with the placenta TFBS clusters , the Z-score for the overlap is more significant . The difference in overlap and Z-score is likely due to the difference in peak width between the datasets: the peaks in the mouse experimental set are 3 , 000 bp wide whereas the peaks in the human experimental set are , on average , 125 bp wide . To check this , we padded each peak in the human experimental data set such that the peak widths were 3 , 000 bp ( with the full set now covering 9 . 26% of the genome ) . This increased the overlap with the placenta TFBS clusters to 796 ( 43% ) , and brought the Z-score ( 53 . 61 ) closer to the Z-score of the mouse experimental data set . We next wanted to check that the placenta TFBS clusters that do not overlap with the mouse and human experimental data are likely to be involved in placenta development . We first identified TFBS clusters that do not overlap with either the mouse or the human experimental data . 831 ( 45% ) of the placenta TFBS clusters that are not within 1 kb of a TSS do not overlap with either experimental dataset ( Figure 5E ) . Additionally , of the candidates that were tested for enhancer activity in TSCs and TGCs , 8 are unique to the placenta TFBS clusters , and 7 of these had more than 2-fold activity compared to the empty vector in at least one of TSCs and TGCs . GREAT analysis of the 831 elements unique to placenta TFBS clusters shows they still have strong enrichment for placenta functions; for example , “abnormal trophoblast layer morphology” has a q-value of 2 . 49×10−8 and a fold enrichment of 3 . 53 ( Supplementary Table S6 ) . We next determined if the fraction of regions associated with a placenta term in GREAT was higher for unique placenta TFBS clusters , compared to regions that were only identified in the mouse experimental set , or regions that were only identified in the human experimental enhancer set . We found that the unique placenta TFBS clusters have between 1 . 21-fold and 2 . 60-fold more regions associated with placenta terms ( Table 2 ) . The GREAT terms are somewhat incomplete , in that every gene involved in placenta development has not been characterized . Nevertheless , this test suggests that if choosing a candidate randomly from the sets that are unique to each method , a candidate chosen from the placenta TFBS clusters is more likely to function in the placenta . Because our pipeline filters for regions of the genome that contain an abnormally high number of non-placenta TFBS clusters , we also compared our data to a filtered version of the mouse experimental data , containing putative placenta-specific enhancers . This set contains 4 , 326 regions ( >1 kb from a gene TSS ) and was generated using a tissue-specificity index based on H3K4me1 occupancy in the 18 tissue and cell types described above [4] . GREAT analysis shows that while the putative placenta-specific enhancers from [4] are more enriched for placenta terms than the full set of mouse experimental data , 5 out of 6 placenta terms have higher fold enrichment in GREAT analysis of our placenta TFBS clusters ( Supplementary Table S7 ) . There are only 12 regions shared between the placenta TFBS clusters and mouse putative placenta-specific enhancers , and for the same 5 out of 6 placenta terms , regions unique to the placenta TFBS clusters have a higher fraction of regions associated with the placenta terms ( Supplementary Table S7 ) . Finally , we compared our data to recently published ChIP-Seq data for enhancer marks and repressor marks in mouse TSCs [22] . One would hope that our placenta TFBS clusters have significant overlap with enhancer mark ChIP-Seq data , and little overlap with repressor mark ChIP-Seq data in this cell type , and this is indeed the case . While 299 placenta TFBS clusters overlap with TSC ChIP-Seq enhancer-associated peaks ( enrichment p<10−4 , Z-score = 21 . 93 ) , only 1 placenta TFBS cluster overlaps with TSC ChIP-Seq repressor associated peaks ( H3K9me3: depletion p<10−4 , Z-score = 8 . 09; H3K27me3: depletion p<10−4 , Z-score = 4 . 13 ) ( Supplementary Figure S9 ) . Together , these data show that while the large-scale experimental data are valuable , they can be strengthened with complementary computational analysis . The method we describe provides a smaller , more focused set with additional candidates that are near genes functioning in placenta development , and that very likely drive activity in placental cell types . To demonstrate that our approach can be generalized , we applied our method to identify TFs and TFBS clusters in five additional tissues: heart , pancreas , blood vessel , bone marrow , and liver . The top 50 motifs and all TFBS clusters for each tissue are provided as Supplementary Tables S8 , S9 , S10 , S11 , S12 . We first examined whether TFs annotated by GREAT as contributing to tissue development are indeed enriched in our predicted top 50 TFs per tissue ( Supplementary Figure S10 ) . The enrichment is strong in four of five tissues ( heart , blood vessel , bone marrow and liver; all p<10−3 ) , but not in pancreas ( p = 0 . 1 ) . When we examine the top GREAT enrichment for the TFBS clusters for each tissue we obtain a similar picture: clusters for the same four tissues yield a top prediction that matches the tissue identity ( e . g . “abnormal ventricle myocardium morphology” for heart ) , while the top term for pancreas is a mismatched brain related term ( Table 3 ) . Because our method identifies the most relevant TFs for a tissue based on enrichment of the TFBS near target genes already annotated to have a role in the tissue , we expected that our approach would be most suitable for tissues with well-annotated terms . If a GREAT term is not well annotated for a tissue , then it is more difficult to determine if the binding site predictions for a particular TF are enriched near genes involved in development of the tissue , because the number of genes that are known to be relevant for the tissue is low . Indeed , we see that the GREAT term used to identify TFs involved in pancreas development is only annotated with 75 target genes , compared to 340–1 , 047 genes associated with terms used for the other tissues investigated ( Supplementary Table S13 ) . These data confirm that our approach works best when genes involved in the tissues or processes are well annotated , likely because of increased ability to predict the TFs that are the most relevant to the tissue .
Here we describe a novel method to identify previously uncharacterized TFs that may have a role in a tissue of interest , as well as active enhancers particularly relevant to the tissue of interest . The automated identification of TFs relevant to a tissue overcomes limitations of current methods for computational identification of enhancers . We first used our method to implicate 50 TFs in placenta development , 33 of which were confirmed to have roles in placenta development in the literature . We predict that the 17 remaining TFs have a role in placenta development , and binding sites for these TFs would not be included as input for previous computational methods that rely heavily on manual curation of TFs . There are multiple lines of evidence supporting our prediction that the remaining 17 TFs have a role in placenta development , both through our analysis and the literature . For example , the highest ranking TF ( 4th ) we predict to have a role in placenta development is Zbtb7b . Zbtb7b knockout mice have defects in the hematopoietic system , and defects in T-cell development and differentiation [33] . Furthermore , intercrosses of mice with a mutation in Zbtb7b produce small litters , whereas wild-type females crossed with mutant males have normal litter sizes [33] . This suggests impairment of female fertility in mutant mice , perhaps due to defects in the decidua , the maternal component of the placenta . ZBTB7B is also highly expressed in three components of the human placenta , including the decidua ( top quartile for expression value ) [21]; is highly expressed in mouse TSCs [22]; and has been shown to be up-regulated in placentas from pregnancies that resulted in intrauterine growth restriction ( IUGR ) , or placentas from pregnancies that resulted in preeclampsia and IUGR [34] . To further investigate and confirm the importance of the TFs we predict to be involved in placenta development , placenta-specific mouse misregulation models could be generated [35] , [36] . We next used our method to identify placenta enhancers by searching for regions of the genome containing clusters of five or more non-overlapping placenta TFBS . We found that the set of placenta TFBS clusters was highly enriched for placenta terms in GREAT , and we generated a null model to ensure the enrichment was due to choosing regions with ≥5 placenta TFBS . We validated our approach using luciferase reporter assays for two placental cell types: TSCs , and TGCs differentiated from TSCs . 31 ( 86% ) of the candidates we tested were active in at least one of TSCs and TGCs . It remains possible that the 5 ( 14% ) that did not show activity are active in a different placenta cell type . Of the 5 candidates that do not show activity in TSCs or TGCs , 4 ( 80% ) are annotated as biochemically active in the mouse or human experimental placenta datasets [2]–[4] . Many of the candidates we tested are in the regulatory domain of genes that are well studied and have a known role in placenta development . The enhancers we identify near these genes have not been characterized , and could be important regulators of the placenta genes . We also show that of the 19 placenta TFBS clusters we tested that consist of more binding sites for TFs we predict to have a role in placenta development , 16 ( 84% ) are active in at least one of TSCs and TGCs . These enhancers would not be identified in computational screens that first identify relevant TFs based on manual curation and then search for clusters of binding sites for only those TFs . We also showed that we can add value to the experimental assays used to generate large-scale data sets through the ENCODE and Roadmap Epigenomics projects . These projects have provided tens of thousands of regions , consisting of enhancer-associated chromatin marks and open chromatin that may be functional during a specific time of placenta development . Our approach allows focusing on regions that are likely more specific for placenta functionality , and identifies many additional putative placenta enhancers , unique to those identified in the experimental datasets . The comparison between putative enhancers identified through our computational approach and putative enhancers identified through epigenomic approaches demonstrates that both approaches likely result in numerous false negatives . While our approach identifies putative enhancers that are missed at specific time-points assayed , we only capture a subset of the putative enhancers identified through experimental approaches . Therefore , it is the combination of computational and experimental approaches that will allow us to more comprehensively understand the enhancers that govern embryo development , through its many tissue and time point combinations . To show that our approach can be generalized , we applied it to generate TFBS clusters in 5 other tissues . Of these tissues , 4 yielded a relevant top enrichment in GREAT . The tissue for which TFBS clusters did not result in a relevant GREAT enrichment was limited by the low number of genes annotated for the relevant GREAT term . These results show that the computational framework described can be easily adapted to other tissues , developmental processes , or across different environmental conditions for which functional annotations are available . This can help overcome the burden associated with carrying out biochemical assays on every tissue and experimental condition and will become even more powerful as gene ontologies for various tissues and processes continue to improve . Gene ontologies are also becoming more specific , and terms in the ontologies more often relate to a particular cell type of a tissue . Tissues are generally not made up of homogenous cell populations , so as the more specific terms become better annotated , our approach is expected to provide enhancer sets for these different cell types . Comparison of mouse and human placentas has shown that gene expression patterns and pathways are often conserved [37] , [38] . The conserved placenta TFBS clusters we identify are likely functional in both species . Using a conservation metric provides confidence in the functionality of the elements we identify , as it has been shown that conserved binding sites are more likely to lie within active enhancers [39] , [40] . Additionally , our method of identifying commonalities between mouse and human is beneficial as it allows us to learn about the human condition by using insights from the mouse model . Transcriptional regulation is a complex process , and identification of enhancers that regulate each developmental process is a challenging task . We have shown that the method we described can be used to accurately predict enhancers in the placenta . This method can be generalized to other tissues , and can complement tissue and time-point restricted data coming from projects such as ENCODE and the Epigenomics Roadmap , highlighting enhancers that have been under purifying selection through mammalian evolution , and therefore are more likely to contribute to phenotypic and disease susceptibility differences .
A transcription factor motif library was curated as described previously [7] , resulting in a non-redundant set of 917 motifs from UniPROBE [12] , JASPAR [13] , and TransFac [14] . The excess conservation method used for binding site prediction was described previously [7] . Binding site predictions for our motif library were carried out in mouse mm9 , ensuring that all predictions were conserved in human hg18 , using the PRISM pipeline for binding site prediction and scoring with the following parameters: binding site prediction threshold was 800 , binding sites were allowed to shift by 20 base pairs relative to the reference , binding sites were required to have a minimum branch length of 2 substitutions per site , the binding site had to be present in at least 5 species , one of those species had to be human , and the p-value of the observed motif score against motif shuffles in similarly conserved windows had to be ≤0 . 05 . The top 10 , 000 predictions , ranked by p-value , were obtained for all 917 motifs . If a motif did not have 10 , 000 predictions satisfying the prediction criteria , then the lower number of predictions for that motif was used . If the motif had additional predictions after the top 10 , 000 with the same score ( ties ) as the last prediction included , they were also included for further analysis . Each set of predictions was run through GREAT [8] using default parameters and a binomial fold enrichment cutoff ≥1 . 5 . GREAT results were filtered for “placenta” and “trophoblast” , using only GO Biological Process , MGI Phenotype , and MGI Phenotype Single Knockout ( KO ) ontologies . The MGI Phenotype Single KO ontology is a version of the MGI Phenotype ontology that only includes single mutant gene to phenotype associations . Each motif was assigned a q-value , based on the best q-value for all placenta terms . Motifs were then sorted by these q-values . Similar Position Weight Matricies ( PWMs ) were grouped in order to remove redundant binding site predictions as described previously [41] . The similarity of two motifs was defined as the maximum pairwise alignment score achieved when all alignments of the two motifs were assessed by shifting the motifs relative to each other for both orientations of the motifs . The alignment score was defined as the sum of column scores for all aligning columns normalized by the geometric mean of the self-alignment of each motif to itself . The column scoring function used was: The column score function attempts to capture the probability that a specific column will select the same base and is then penalized by the chance that it will not select the same base . A similarity threshold of 0 . 85 was used . To generate a list of known placenta TFs using an automated approach , we combined gene lists from GREAT for the most general placenta terms , abnormal placenta morphology ( GO Biological Process ontology ) , and placenta development ( Mouse Phenotype Single KO ontology ) . This resulted in a list of 349 genes . We then associated PWMs with gene names to classify each group of similar PWMs as ‘known’ or ‘unknown’ depending on whether any single PWM in a group of similar motifs mapped to the list of 349 genes . After each motif was classified using this method , we determined whether a significant number of TFs in the top 50 appeared in the known placenta gene list . We did so by comparing the number of ‘known’ TFs in the top 50 to the number of TFs that appear in the known placenta gene list when 50 random TFs ( below rank 100 ) were chosen from the list , a total of 1 , 000 times . We also calculated a Wilcoxon rank-sum p-value to determine if known TFs were enriched toward the top of the list . For each of the other tissues analyzed , we generated a list of known TFs using the same approach , but with the following GREAT terms: heart: heart development , abnormal heart development; pancreas: pancreas development , abnormal pancreas development; blood vessel: blood vessel development , abnormal blood vessel morphology; bone marrow: abnormal bone marrow cell morphology/development; liver: liver development , abnormal liver size . To more carefully determine if the top 50 TFs in our list were known to be involved in placenta development , we searched the literature . For a TF to be considered ‘known’ or ‘well-studied’ in the placenta , the literature must show strong experimental evidence , such as placenta abnormalities upon gene knockout , or defects in trophoblast function upon gene knockout/knockdown/overexpression in relevant placenta cell lines . TFs that have structurally similar family members involved in placenta development for which we do not have motifs were also considered ‘known’ . Additionally , TFs with PWMs that closely resemble PWMs that map to genes involved in placenta development , but were just below the 0 . 85 grouping threshold , were considered ‘known’ . For a TF to be considered ‘predicted’ , single studies may have implicated the TF in placenta development , but the relationship has not been well characterized . Additionally , if the only evidence for a TF's involvement in placenta was gene expression , the TF was considered ‘predicted’ . To identify potential placenta enhancers , we used a hierarchical clustering approach ( UPGMA ) over the genome to search for binding site predictions in close proximity of each other for the 50 TFs described above . We first placed each binding site prediction in its own TFBS cluster , mapped to its own centroid . We then iteratively agglomerated the two TFBS clusters with the nearest centroids . Each TFBS cluster was scored based on the number of non-overlapping binding site predictions ( TFBS that share ≤3 bp ) falling within it . To reward TFBS density , the scores for regions longer than 250 bp were weighted by a penalty function: exp ( −0 . 5 ( regionLength−250 ) 2/2502 ) . TFBS clusters were ranked by score , and then the ranked list was traversed , outputting only those TFBS clusters that did not overlap a previously output TFBS cluster . We discarded TFBS clusters that overlapped with exons . GREAT analysis was performed using default GREAT filters for significant terms: region-based fold enrichment ≥2 and false discovery rate ( FDR ) q-value ≤0 . 05 , with the additional requirement that at least 25 genes in the term were hit . Unless specifically noted , GREAT results were filtered by fold enrichment . TSCs ( a kind gift from Dr . Emin Maltepe at UCSF ) were grown , differentiated , and passaged according to standard protocols [42] . Passage 2 mouse embryonic fibroblasts ( MEFs ) ( Applied Stem Cell ) were expanded and treated with Mytomycin C as previously described [42] , aliquoted and frozen to use as feeder cells for TSCs . TSCs were split once a week at a 1∶50 dilution onto a plate of fresh MEFs . For differentiation into trophoblast giant cells ( TGCs ) , a 1∶10 dilution of confluent TSCs was plated onto a 10 cm plate . Five days later , the differentiating TSCs were split into a 24-well plate at a 1∶8 dilution . Inserts were amplified from mouse genomic DNA ( Clonetech Laboratories , Inc . ) using Phusion High Fidelity DNA Polymerase ( NEB , Inc . ) and cloned into the 5′ KpnI and 3′ HindIII sites of pGL4 . 23 ( Promega , Corp . ) . A second reporter vector was constructed , pGL4 . 23 LIC , by introducing a Ligation Independent Cloning ( LIC ) linker into the 5′KpnI and 3′ HindIII sites of pGL4 . 23 . The LIC forward site was: 5′-cGCTCTTCGGGATGGAGGGATATCCACCTTACCCGAAGAGCa-3′ and the LIC reverse site was: 5′-agcttGCTCTTCGGGTAAGGTGGATATCCCTCCATCCCGAAGAGCggtac-3′ . The genomic inserts were cloned into the pGL4 . 23 LIC vector using an LIC method described previously [43] . All positive clones were identified by colony PCR and sequenced . Primers used to amplify genomic regions are listed in Supplementary Table S4 . Transfections were done according to the Invitrogen protocol for Lipofectamine LTX & Plus reagent . For TSCs , a confluent 10 cm plate was split 1∶4 into a 24 well plate . We used a 1 µg∶4 µl ratio of DNA to reagent , and transfected 1 µg of reporter construct and 20 ng of pRL-TK vector ( used as a transfection efficiency control vector , Promega Corp . ) per well . Cells were lysed 24 hours post-transfection and frozen until luciferase assays were performed . For TGCs , plates were transfected 12 days after starting differentiation . We used a 1 µg∶3 µl ratio of DNA to reagent , and transfected 750 ng reporter construct and 15 ng of pRL-TK vector ( used as a control vector , Promega Corp . ) . Cells were lysed 48 hours post-transfection and frozen until luciferase assays were performed . Each candidate was tested in triplicate within a single plate ( technical replicates ) , and on at least 3 different days ( biological replicates ) . Primary neocortical cells were prepared from e14 . 5 mice and transfected with pGL4 . 23 or pGL4 . 23 LIC containing the genomic regions and pRL-CMV ( as a control vector ) using the Amaxa 96-well Shuttle Protocol for Primary Mammalian Neurons ( Lonza ) [25] . We used 2 . 5×105 cells , 100 ng plasmid DNA , and 60 ng pRL-CMV per nucleofection sample . Transfected cells were resuspended in 120 µl supplemented PNBM media ( Lonza ) and plated on a 96-well plate treated with Poly-D-Lysine . 40 µl of this suspension were added to 160 µl PNBM per well . Cells were lysed 48 hours post-transfection and frozen until luciferase assays were performed . Luciferase assays were done using the DLR kit ( Promega ) according to manufacturer's instructions and read using a Promega Glomax luminometer using the “Dual-Luciferase 2 injectors” program with a 50 µl injection volume for both LAR II and Stop & Glo Reagent . Mouse placenta data from [4] was downloaded from the Ren Lab website: http://chromosome . sdsc . edu/mouse/download . html We took the union of the regions in the placenta . enhancer . txt file with the regions in the placenta . h3k27ac . peak . txt file , after padding each given coordinate by ±1500 bp , as recommended by the Ren lab . We then removed regions within 1 kb of gene transcriptional start sites , resulting in 70 , 951 peaks . Placenta-specific regions were downloaded from the same website , and were similarly padded by ±1500 bp . Mouse tissue data from [4] was downloaded from the Ren Lab website: http://chromosome . sdsc . edu/mouse/download . html We downloaded enhancer files for 18 tissues and cell types , and padded each given coordinate by ±1500 bp before determining the overlap with the placenta TFBS clusters . We analyzed the following tissues and cell types: cortex , MEFs , bone marrow , cerebellum , e14 . 5 liver , e14 . 5 brain , e14 . 5 heart , e14 . 5 limb , liver , heart , intestine , kidney , spleen , lung , mESCs , olfactory bulb , testes , and thymus . DNAse I hypersensitive sites in human placenta were provided by the Stamatoyannopoulos lab [2] , [3] . Data from 6 samples were provided , aged at 113 days gestation , 108 days gestation , 105 days gestation , 91 days gestation , and two at 85 days gestation . Peaks from samples below 100 days old were intersected and peaks from replicates above 100 days old were intersected . The union of the two sets was then taken and converted to mm9 coordinates using UCSC's liftover tool with default parameters . We then removed regions within 1 kb of gene transcriptional start sites , resulting in 80 , 922 peaks . TSC data are from [22] and were downloaded from GEO . RNA-Seq data were ranked according to average tag count between biological replicates , normalized to the 3′ UTR length for reported genes . TSC H3K27ac and H3K4me1 data were intersected to generate a TSC putative enhancer set . For each set of peaks we used ( H3K27ac , H3K4me1 , H3K27me3 , H3K9me3 ) , regions within 1 kb of gene TSS were removed . | Enhancers are distal gene regulatory elements that can activate tissue- and time-point specific gene expression . Identification of active enhancers is challenging , and is the subject of intense investigation . We developed an automated computational framework to predict transcription factors ( TFs ) and enhancers that target a tissue of interest by combining two growing resources: TF binding motifs and target gene function annotations . We applied our framework to the placenta , and confirmed our enhancer predictions are more active in placental cell types than others . To demonstrate generalizability , we applied our approach to 5 additional tissues . The combination of experimental sampling with computational prediction approaches will aid in the identification of those enhancers that are most likely active in a particular tissue , as well as the characterization of groups of TFs associated with these enhancers . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genomics",
"functional",
"genomics",
"biology",
"computational",
"biology"
] | 2014 | Automated Discovery of Tissue-Targeting Enhancers and Transcription Factors from Binding Motif and Gene Function Data |
Zika virus ( ZIKV ) has emerged since 2013 as a significant global human health threat following outbreaks in the Pacific Islands and rapid spread throughout South and Central America . Severe congenital and neurological sequelae have been linked to ZIKV infections . Assessing the ability of common mosquito species to transmit ZIKV and characterizing variation in mosquito transmission of different ZIKV strains is important for estimating regional outbreak potential and for prioritizing local mosquito control strategies for Aedes and Culex species . In this study , we evaluated the laboratory vector competence of Aedes aegypti , Culex quinquefasciatus , and Culex tarsalis that originated in areas of California where ZIKV cases in travelers since 2015 were frequent . We compared infection , dissemination , and transmission rates by measuring ZIKV RNA levels in cohorts of mosquitoes that ingested blood meals from type I interferon-deficient mice infected with either a Puerto Rican ZIKV strain from 2015 ( PR15 ) , a Brazilian ZIKV strain from 2015 ( BR15 ) , or an ancestral Asian-lineage Malaysian ZIKV strain from 1966 ( MA66 ) . With PR15 , Cx . quinquefasciatus was refractory to infection ( 0% , N = 42 ) and Cx . tarsalis was infected at 4% ( N = 46 ) . No ZIKV RNA was detected in saliva from either Culex species 14 or 21 days post feeding ( dpf ) . In contrast , Ae . aegypti developed infection rates of 85% ( PR15; N = 46 ) , 90% ( BR15; N = 20 ) , and 81% ( MA66; N = 85 ) 14 or 15 dpf . Although MA66-infected Ae . aegypti showed higher levels of ZIKV RNA in mosquito bodies and legs , transmission rates were not significantly different across virus strains ( P = 0 . 13 , Fisher’s exact test ) . To confirm infectivity and measure the transmitted ZIKV dose , we enumerated infectious ZIKV in Ae . aegypti saliva using Vero cell plaque assays . The expectorated plaque forming units PFU varied by viral strain: MA66-infected expectorated 13±4 PFU ( mean±SE , N = 13 ) compared to 29±6 PFU for PR15-infected ( N = 13 ) and 35±8 PFU for BR15-infected ( N = 6; ANOVA , df = 2 , F = 3 . 8 , P = 0 . 035 ) . These laboratory vector competence results support an emerging consensus that Cx . tarsalis and Cx . quinquefasciatus are not vectors of ZIKV . These results also indicate that Ae . aegypti from California are efficient laboratory vectors of ancestral and contemporary Asian lineage ZIKV .
Zika virus ( ZIKV ) is a mosquito-transmitted flavivirus that was first isolated in 1947 in the Zika forest of Uganda from a sentinel rhesus macaque [1] . Since its discovery , human ZIKV cases have been reported across Africa and Asia , but until 2007 the virus received little attention from researchers as it was thought to cause only mild disease . Following epidemics in Micronesia in 2007 , French Polynesia in 2013 , and Brazil in 2015 [2] , ZIKV has now been confirmed as a cause of the neurological disease Guillain-Barre syndrome and congenital disorders , including microcephaly in infants [3] . Despite a dramatic decline in Brazilian cases since 2016 , ZIKV remains a significant global human health threat [4] , as other countries including Argentina , Bolivia , Peru , and Ecuador reported an increase in cases in 2017 [5] . Reducing mosquito vector populations is an effective way to mitigate mosquito-borne disease transmission [6] . Therefore , identifying ZIKV vector species is crucial for accurate risk assessments for mosquito transmission and to target vector control measures to mitigate ZIKV disease . Several Aedes species have been identified as competent vectors in laboratory studies , including the primary vector Aedes ( Ae . ) aegypti [7–21] , Ae . albopictus [7 , 8 , 10 , 11 , 15 , 17 , 19 , 20 , 22 , 23] , Ae . notoscriptus [10] , Ae . camptorhynchus [10] , Ae . luteocephalus [24] , Ae . vexans [25] , and Ae . vittatus [24] . Culex species generally do not become infected with ZIKV and are incapable of transmitting [7 , 9 , 10 , 12 , 14 , 17 , 23 , 26–29] . Exceptions include a study from Guadalajara , Mexico , where infectious ZIKV was detected in pooled mosquito tissue samples from field-collected Cx . tarsalis , Cx . coronator , and Cx . quinquefasciatus [30] . ZIKV RNA has also been detected in pooled field samples of Cx . quinquefasciatus from China [31] . Evidence for ZIKV transmission by Culex species is limited to Cx . quinquefasciatus and includes ZIKV RNA detected in saliva on Flinders Technology Associates ( FTA ) cards provided to a cohort of laboratory-infected mosquitoes from Brazil [32] and transmission to 1-day-old mice from mosquitoes from China , although inconsistently with other murine studies [33–35] , no murine fatality was noted [36] . Previous studies demonstrate that ZIKV vector competence is more complex than simple mosquito species-level designations , and thus region-specific mosquito genotypes and multiple ZIKV strains must be evaluated to assess region-specific vector competence . For example , Ae . aegypti from the Dominican Republic transmit ZIKV isolated from Cambodia in 2010 ( FSS 13025 ) and Mexico in 2015 ( MEX1-7 ) more effectively than Ae . aegypti from Salvador , Brazil [21] . Furthermore , ZIKV from Brazil in 2015 ( BeH815744 ) has higher infectivity than a French Polynesian strain from 2013 ( H/PF13 ) in Ae . aegypti from Singapore [8] . The source of virus also matters; fresh ZIKV was more infectious in comparative studies than freeze-thawed virus [12] . California ( CA ) vector control districts have been combating stable Ae . aegypti populations in the state since 2013 [37] , including in many counties in Southern CA . In addition , between 2015 and March 2018 , 640 travel-associated ZIKV infections were reported in CA [38] , 137 ( 21% of cases in state ) of which were in Los Angeles County where the Ae . aegypti used for vector competence experiments here were collected . Due to the presence of Ae . aegypti and numerous travel-associated ZIKV infections , there is a risk of the establishment of local mosquito-borne ZIKV transmission . Additionally , genetic variation between Central Valley and Southern CA Ae . aegypti populations has been observed [39] , even between populations in neighboring cities , such as Fresno and Clovis [40] . These findings indicate that gene flow is limited between Ae . aegypti populations and leave open the possibility that important traits , such as vector competence , may also vary among Ae . aegypti throughout the state . To better assess local ZIKV transmission risk , we evaluated the laboratory vector competence of Ae . aegypti from Los Angeles , CA , for ZIKV isolates from Puerto Rico ( 2015 ) , Brazil ( 2015 ) , and Malaysia ( 1966 ) . We also evaluated the laboratory vector competence of two highly abundant Culex species , Cx . quinquefasciatus from Orange County , CA , and Cx . tarsalis from Kern County , CA , with a Puerto Rico ( 2015 ) ZIKV strain .
Three Asian-lineage strains of ZIKV were used in our experiments . A 2015 Puerto Rican strain was isolated from human serum in 2015 ( PR15 , PRVABC59 ) , passaged 4 times in Vero cells , and sequenced . The coding sequence for the complete genome of the passaged we used was identical to GenBank accession number KX601168 . An Asian-lineage Malaysian ZIKV strain isolated from Ae . aegypti mosquitoes in 1966 ( MA66 , P6-740 [41] ) that had been passaged in suckling mouse brains 6 times and once in Vero cells before it was received from the Centers for Disease Control was passaged once more in Vero cells . The complete coding genome sequence of our passage of MA66 was 100% identical to GenBank accession number KX601167 . 1 . A Brazilian strain isolated from human serum in 2015 ( BR15 , SPH2015 ) was passaged 3 times in Vero cells and sequenced . The complete genome coding sequence of BR15 was identical to GenBank accession number KU321639 . Strains MA66 and PR15 were obtained from Dr . Aaron Brault at the U . S . Centers for Disease Control and Prevention in Fort Collins , Colorado . Dr . Mike Busch at Blood Systems Research Institute , San Francisco , CA , provided the BR15 strain . All ZIKV strains and their source Vero cells were confirmed mycoplasma negative by PCR according to the manufacturer’s instructions ( Agilent Mycoplasma Plus PCR Primer Kit , Santa Rosa , CA . ) The Ae . aegypti mosquitoes used in this study were field-collected as larvae in Los Angeles , CA , in 2016 and morphologically identified . The F6 generation was used for this study . Adult Cx . quinquefasciatus mosquitoes were field-collected as adults in Orange County , California in 2016 and morphologically identified . The F5 generation was used for this study . The Cx . tarsalis mosquitoes were field-collected in the Kern National Wildlife Refuge , Kern County , CA in 2002 , morphologically identified , and have been maintained continuously in colony since . Female interferon-deficient ( IFN-α/βR−/−; C57BL/6 ) mice aged 4–8 weeks ( B6 . 129S2-Ifnar1tm1Agt/Mmjax , The Jackson Laboratory , Sacramento , CA ) were used for all experiments . Differences in ZIKV viremia levels and kinetics in male versus female mice have not been observed [33] . Mice were inoculated with 5 log10 Vero plaque forming units ( PFU ) of ZIKV via subcutaneous injection . ZIKV-infected mice were presented to mosquitoes 2 days post-inoculation , at peak viremia [33] . Mice were anesthetized prior to mosquito exposure with a ketamine ( VETone Zetamine CIII , 75 mg/kg ) , xylazine ( AnaSed , 10 mg/kg ) , and acepromazine ( AceproJect , 1 mg/kg ) solution administered intraperitoneally . The ZIKV viremia in each mouse was determined by Vero cell plaque assay from 30 μL of whole blood collected immediately prior to the mosquito feed . Viremic mice were presented for two cohorts of adult female mosquitoes 30–60 minutes on one of three arrangements depending on species: ( 1 ) 25 Cx . tarsalis in pint cartons ( amazon . com ) , ( 2 ) 50 Ae . aegypti in pint cartons , or ( 3 ) >100 Cx . quinquefasciatus in a 1 ft3 mesh cage ( BugDorm , MegaView Science , Taiwan ) . Engorged females were sorted from non-fed individuals by vacuum aspiration . Mosquito ages at the time of blood-feeding were 4–14 days post eclosion ( dpe ) for Cx . tarsalis , 14–21 dpe for Cx . quinquefasciatus , and 4–12 dpe for Ae . aegypti . Cx . tarsalis and Ae . aegypti were held at 26°C , 80% relative humidity , and 12:12 h light:dark cycle . Cx . quinquefasciatus were maintained at room temperature ( 22°C and 33% relative humidity ) to ensure survival . All mosquitoes had constant access to 10% sucrose before and after blood-feeding , except during a 24-hour starvation period prior to presentation of the viremic mice . At days 14 and 21 post bloodfeed , mosquitoes were cold-anesthetized at -20°C for 5 minutes and then legs and wings were removed with forceps while immobilized on ice . Saliva was collected by inserting the proboscis into a capillary tube containing fetal bovine serum ( FBS , GenClone ) for 20 minutes . Individual bodies , legs+wings , and the saliva sample from each mosquito were stored separately in 2 mL tubes containing a 5 mm glass bead and 250 μL Dulbecco’s modified eagle medium ( DMEM , Gibco ) supplemented with 50μg/mL of penicillin/streptomycin and 20% FBS . All samples were stored at -80°C until further processing . Mosquito tissues and glass capillary tubes containing saliva samples were homogenized in DMEM by shaking for 2 minutes at 30 shakes/second using a Tissuelyser ( Qiagen , Hilden , Germany ) . Viral RNA was extracted using the MaxMax Viral RNA Extraction Kit ( ThermoFisher , Waltham , MA ) . A total of 50 μL of homogenate for mosquito tissue and 100 μL of saliva samples were extracted . All RNA extracts were eluted in 50 μL of elution buffer ( Buffer EB , Qiagen ) and stored at -80°C until further testing . ZIKV RNA titers were determined for each body , legs+wings , and saliva sample using the Taqman Fast Virus One-Step Master Mix ( ThermoFisher ) reverse transcription RT-qPCR kit with a previously described ZIKV-specific assay ( primers: ZIKV 1086 , ZIKV 1162c , and ZIKV 1107-FAM; [42] ) . At least two technical replicates were performed for all samples . Samples with a mean cycle threshold ( Ct ) value of 38 or below were considered positive for ZIKV RNA . This limit of detection was determined from prior testing of serially diluted samples of known ZIKV RNA concentrations with the same extraction and RT-qPCR reagents and protocols and equipment [43] . To estimate infectious ZIKV in expectorated Ae . aegypti saliva , viral titrations were performed on a random sample of RT-qPCR-positive saliva samples at the second or third thaw in Vero cell culture by plaque assay . In brief , cell monolayers were inoculated with 110 μL of undilute saliva from individual mosquitoes mixed with DMEM containing 2% ( vol/vol ) FBS , and 100 U/mL penicillin/streptomycin . After a one hour incubation period to allow for viral infection of cells , 0 . 8% agarose/DMEM was added to cover the cells . The plates were incubated at 37°C in 5% CO2 for 8 days . The cells were then fixed with 4% formaldehyde and stained with 0 . 05% crystal violet . Plaques were visualized as holes in the Vero cell monolayer and counted to determine PFU values . The limit of detection of the assay was 2 . 3 PFU where 110 μL of the total saliva sample ( 250 μL ) was inoculated directly onto the cells . Since the volume of saliva was limited , each sample was tested in just 1 replicate . In this study , we calculated infection rates as the number of RT-qPCR positive individual bodies divided by the number of individuals that ingested blood and were tested , dissemination rates as the number of RT-qPCR positive pooled leg & wing sets from each individual divided by the number of individuals that ingested blood and were tested , and transmission rates as the number of RT-qPCR positive saliva samples divided by the total number of individuals that ingested blood and were tested . For Ae . aegypti and Cx . tarsalis , multiple cohorts of the same species fed on different mice infected with the same ZIKV strain with slight ( ≤1 log10 ) variations in viremias . Preliminary analysis across same-species cohorts that fed on different mice infected with the same ZIKV strain revealed no significant differences ( Fisher’s exact test , P>0 . 05 ) in infection , dissemination and transmission rates . We therefore combined the data presented for each ZIKV strain for Ae . aegypti and Cx . tarsalis , while also reporting the magnitudes of viremia in all mice ( Table 1 ) . Comparisons of ZIKV RNA levels and PFU in saliva samples between ZIKV strains was performed using a one-way ANOVA with Tukey’s correction for pairwise comparisons ( reported as Padj ) and ZIKV RNA detection rates were compared using two-tailed Fisher’s exact tests ( scipy . stats ) . Data were plotted using matplotlib ( Python ) . All procedures involving mice were performed in accordance with IACUC protocol #19404 that was reviewed and approved by the UC Davis IACUC on June 29 , 2017 . The UC Davis IACUC adheres to the Office of Laboratory Animal Welfare Health Research Extension Act of 1985 ( Public Law 99–158 ) as well as the United State Department of Agriculture’s Animal Welfare Act . UC Davis is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care , International ( AAALAC ) and has an Animal Welfare Assurance ( number A3433-01 ) on file with the Office of Laboratory Animal Welfare ( OLAW ) .
Cx . tarsalis and Cx . quinquefasciatus mosquitoes were tested 14 or 21 days after ingesting blood from ZIKV-infected interferon receptor deficient mice . Two Cx . tarsalis bodies out of the 46 individuals tested ( 4% ) had low levels of ZIKV RNA at 14 dpf ( Ct < 38; 48 ZIKV genomes/body ) . Both infected individuals also had detectable ZIKV in their legs and wings , indicating disseminated infections . Neither of the ZIKV-infected Cx . tarsalis contained detectable ZIKV RNA in their saliva samples ( Table 1 ) . The Cx . tarsalis infection rate significantly increased from 4% to 30% ( 2/46 to 6/20 , P<0 . 01 Fisher exact test ) from 14 to 21 dpf . Among the 6 infected Cx . tarsalis at 21 dpf , ZIKV RNA was detected in 1 leg and wing sample but , consistent with a lack of transmission 14 dpf , no ZIKV RNA was detected in the saliva ( Table 1 ) . We did not detect ZIKV RNA in any Cx . quinquefasciatus mosquito tissues 14 ( N = 42 ) or 21 dpf ( N = 37; Table 1 ) . At 14 dpf , ZIKV infection , dissemination , and transmission rates measured by the presence of ZIKV RNA in Ae . aegypti that ingested MA66 were 86% , 79% , and 53% , respectively ( Table 1 ) . For Ae . aegypti that ingested ZIKV PR15 , the infection , dissemination , and transmission rates on 14 dpf were 85% , 78% , and 65% , respectively ( Table 1 ) . ZIKV BR15-exposed individuals harvested 15 dpf had infection , dissemination , and transmission rates of 90% , 90% , and 75% , respectively ( Table 1 ) . ZIKV RNA infection , dissemination , and transmission rates in Ae . aegypti that ingested MA66 or PR15 at 21 dpf were equal or higher than 14 dpf rates . The transmission rate between 14 and 21 dpf increased significantly in Ae . aegypti infected with MA66 ( 53% vs . 87% , P<0 . 01 , Fisher’s exact ) , but not PR15 ( 65% vs . 74% , P = 0 . 59 , Fisher’s exact; Table 1 ) . Transmission rates were not significantly different across viruses ( P = 0 . 13 , Fisher’s exact ) . The mean ZIKV RNA level ( 8 . 9 log10 ) in MA66-infected bodies was significantly higher than the mean for BR15 ( 8 . 2 log10 , ANOVA , degrees of freedom ( df ) = 2 , F-statistic ( F ) = 16 . 3 , Padj<0 . 01 ) and PR15-infected individuals ( 8 . 4 log10 , ANOVA , df = 2 , F = 16 . 3 , Padj<0 . 01; Fig 1 ) . The mean ZIKV RNA level in MA66-infected leg+wing tissue ( 7 . 5 log10 ) was also significantly higher than PR15-infected leg+wing samples ( 7 . 0 log10 , ANOVA , df = 2 , F = 8 . 4 , Padj<0 . 01 ) . Higher ZIKV RNA levels in MA66-infected Ae . aegypti likely do not reflect the dose ingested , where flavivirus infections of mosquitoes typically show a strong dose response , since viremias in both ZIKV MA66-infected mice were lower than those for PR15 . ZIKV RNA levels in saliva were not significantly different among strains ( ANOVA , df = 2 , F = 0 . 96 , P = 0 . 39 ) . A bimodal distribution of ZIKV RNA levels was observed across cohorts of ZIKV PR15- or MA66-infected bodies , with high ( >6 log10 genomes/body ) and low ( <6 log10 genomes/body ) clusters of individuals ( Fig 1 ) . MA66-infected Ae . aegypti that were highly infected ( >6 log10 genomes/body ) had higher transmission rates ( 81% , N = 54 ) compared to low titer ( <6 log10 genomes/body ) individuals ( 5% , N = 19; P<0 . 0001 , Fisher’s exact ) . We also examined the relationship between infection , dissemination and transmission at an individual mosquito level for Ae . aegypti ( Fig 2 ) . Most Ae . aegypti that became infected developed disseminated infections . Individuals with higher ( red/pink in figure ) ZIKV RNA levels in legs+wings were more likely to transmit ZIKV RNA than mosquitoes with low ( blue in figure ) RNA levels in legs+wings . None of the PR15-infected Ae . aegypti with <6 log10 genomes/body transmitted ZIKV RNA ( N = 4 ) . To confirm infectivity and measure the transmitted ZIKV dose , plaque assays were performed on Ae . aegypti saliva collected 14 or 15 dpf to enumerate infectious ZIKV in Vero cell plaque forming units ( PFU ) . Out of 45 RTq-PCR positive saliva samples that were tested by plaque assay , 32 ( 71% ) yielded at least 1 detectable plaque . The expectorated PFU varied by viral strain: the MA66-infected individuals transmitted 13±4 PFU ( mean±SE , N = 13 ) compared to 29±6 for PR15 ( N = 13 ) and 35±8 for BR15 ( N = 6; ANOVA , df = 2 , F = 3 . 8 , P = 0 . 035; Fig 3 ) .
Understanding the mosquito species that vector ZIKV is important for estimating regional outbreak potential and for informing local mosquito control strategies , especially since Aedes and Culex species differ in life history traits and host-seeking behaviors that could impact control efforts . For example , oviposition traps bias towards Ae . aegypti that lay in artificial containers [44] while Culex typically prefer natural pools [45] . For Cx . tarsalis , we detected an overall ZIKV infection rate of 12% ( 8/66 ) in mosquitoes tested 14 and 21 dpf . Disseminated infections in Cx . tarsalis were detected at <5% on both 14 and 21 dpf , with high Ct values indicating low ZIKV RNA levels . We postulate that the disseminated infections detected in Cx . tarsalis may reflect false positives given that mosquitoes with true disseminated infections typically achieve very high viral RNA titers due to prolonged infection of multiple tissues . The absence of detectable ZIKV RNA in saliva at 14 or 21 dpf is evidence that Cx . tarsalis from CA is not capable of transmitting ZIKV in laboratory experiments . Furthermore , Cx . tarsalis feeds less often on human hosts compared to the highly anthropophilic Ae . aegypti [45–47] , making human-mosquito-human ZIKV transmission by Cx . tarsalis unlikely . We also found no evidence for ZIKV infection of Cx . quinquefasciatus from California , with no ZIKV RNA detected in bodies , legs/wings or saliva from nearly 80 individuals . This is the first data showing ZIKV vector competence for California mosquitoes , and it supports results from many other studies which demonstrate that Cx . quinquefasciatus is not a competent laboratory vector of ZIKV . By contrast , Ae . aegypti mosquitoes exhibited infection rates of 85–90% and transmission rates of 53–80% at 14 dpf . The transmitted dose of infectious ZIKV by Californian Ae . aegypti is consistent with the range of doses observed in similar studies with Brazilian Ae . aegypti [48 , 49] . Ae . aegypti that ingested ZIKV MA66 in our laboratory vector competence studies developed higher ZIKV RNA levels than PR15- or BR15-infected mosquitoes . This pattern contrasted with the lower transmission rate and lower expectorated PFU of MA66-infected Ae . aegypti at 14 dpf . A possible explanation for the lower transmissibility of MA66 at 14 dpf is that it lacks a A188V mutation in the NS1 gene that both PR15 and BR15 possess , which has been linked to higher infectivity ( where infectivity can influence transmissibility ) in mosquitoes when interferon-deficient mice are used for blood-feeding [7] . ZIKV strains from recent American outbreaks have also been shown to exhibit higher infection and transmission rates than historic Asian-lineage strains [8] . Additional vector competence studies involving region-specific Ae . aegypti and Ae . albopictus mosquito populations with sequenced genomes and multiple distinct ZIKV isolates will improve our understanding of the both mosquito and virus genetics involved in ZIKV vector competence , which could inform our ability to accurately estimate regional outbreak potential . Among ZIKV MA66-infected Ae . aegypti , we observed that mosquitoes with low RNA copy numbers in bodies were less likely to transmit than those with infections that exceeded 6 log10 genomes per body . This pattern is consistent with the presence of a midgut barrier to infection [50] . In that case , the mosquitoes with low body RNA levels represent infections that have not escaped the midgut while mosquitoes with high body RNA levels correspond to individuals with ZIKV that has disseminated to secondary amplification tissues . This laboratory vector competence study confirmed that Ae . aegypti from Los Angeles , California , USA , can transmit Asian lineage ZIKV and that Cx . tarsalis and Cx . quinquefasciatus are inefficient ZIKV vectors . Given that Culex mosquitoes are poor ZIKV vectors and seek primarily non-human hosts , they are unlikely to facilitate a ZIKV outbreak . Thus , vector control efforts targeting ZIKV should remain focused on reducing urban Aedes populations . | Assessing the ability of common mosquito species to transmit Zika virus ( ZIKV ) and characterizing variation in mosquito transmission of different ZIKV strains is important for estimating regional outbreak potential and for prioritizing local mosquito control strategies for Aedes and Culex species . In this study , we evaluated the laboratory vector competence of Aedes aegypti , Culex quinquefasciatus , and Culex tarsalis that originated in areas of California where ZIKV cases in travelers since 2015 were frequent . We observed variation in infection loads between ZIKV strains in Ae . aegypti , but transmission rates were not different . In addition , there was a positive relationship between ZIKV RNA levels in infected mosquitoes ascertained from bodies and ZIKV RNA transmission rates . Our data add to the growing body of evidence supporting the role of Aedes aegypti as a ZIKV vector and refute Cx . quinquefasciatus and Cx . tarsalis as vectors . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"invertebrates",
"vero",
"cells",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"biological",
"cultures",
"microbiology",
"saliva",
"animals",
"viruses",
"rna",
"viruses",
"mammalian",
"genomics",
... | 2018 | Vector competence of Aedes aegypti, Culex tarsalis, and Culex quinquefasciatus from California for Zika virus |
Expansion or shrinkage of existing tandem repeats ( TRs ) associated with various biological processes has been actively studied in both prokaryotic and eukaryotic genomes , while their origin and biological implications remain mostly unknown . Here we describe various duplications ( de novo TRs ) that occurred in the coding region of a β-lactamase gene , where a conserved structure called the omega loop is encoded . These duplications that occurred under selection using ceftazidime conferred substrate spectrum extension to include the antibiotic . Under selective pressure with one of the original substrates ( amoxicillin ) , a high level of reversion occurred in the mutant β-lactamase genes completing a cycle back to the original substrate spectrum . The de novo TRs coupled with reversion makes a genetic toggling mechanism enabling reversible switching between the two phases of the substrate spectrum of β-lactamases . This toggle exemplifies the effective adaptation of de novo TRs for enhanced bacterial survival . We found pairs of direct repeats that mediated the DNA duplication ( TR formation ) . In addition , we found different duos of sequences that mediated the DNA duplication . These novel elements—that we named SCSs ( same-strand complementary sequences ) —were also found associated with β-lactamase TR mutations from clinical isolates . Both direct repeats and SCSs had a high correlation with TRs in diverse bacterial genomes throughout the major phylogenetic lineages , suggesting that they comprise a fundamental mechanism shaping the bacterial evolution .
As a ubiquitous feature of genomes , tandem repeats ( TRs ) are the sites at which recombination or replication slippage can occur [1]–[3] . Changes in the number of repeat units can confer phenotypic variability in eukaryotes , such as plasticity in skeletal morphology and tuning of the circadian rhythm , and are critical in repeat expansion diseases in humans , such as Huntington's disease [1] . In microorganisms , changes in TRs are the basis for one of the simplest and most prevalent reversible stochastic switching mechanisms , which is commonly known as “phase- or antigenic variation” [2]–[4] . Phase variation generally involves reversible switching that results in an “all-or-none” expressing phase of proteins , whereas antigenic variation alters the surface architecture of proteins that interact with the environment [3] , [4] . Whereas biological consequences affected by alterations in preexisting TRs have been widely reported , processes underlying de novo TR formation and their biological implications have not been actively investigated . Along this line , it is intriguing that there have been reports of duplication mutations that occurred in the coding region of β-lactamase genes , expanding the substrate spectrum of the enzyme to include ceftazidime , a third-generation cephalosporin . A duplication of five amino acids was found in SHV-16 in a clinical isolate of Klebsiella pneumoniae [5] within the omega loop , which is a highly conserved structural domain constituting part of the active-site pocket [6] . In another report , a duplication of three residues was found in the omega loop of a class C β-lactamase in clinical strain Enterobacter cloacae GC1 [7] . The adaptation of β-lactamases in response to exposure to new antibiotics has been a major public health concern , and in almost all cases , point mutations resulting in an amino acid substitution in the enzymes have been responsible for this problem [8]–[10] . Although the biological legitimacy was unclear , the two cases of rare duplication mutations suggested that β-lactamases have potential as an excellent subject to investigate the nature of de novo TRs in connection with the evolution of the drug resistance . In this study , we describe eleven de novo TRs that can occur in the coding region of a β-lactamase gene and allow genetic toggling when coupled with reversion , adjusting the substrate spectrum of the β-lactamase to a different β-lactam antibiotic challenge . At the DNA level , we describe pairs of direct repeats and a novel group of duo elements that we found instrumental in DNA duplication . We then note our findings of a high correlation that exists between direct repeats and TRs and also between the novel duo elements and TRs , supporting the notion that the DNA duplication we described here comprises a fundamental mechanism in bacterial genome evolution .
To determine if TRs can be formed naturally in β-lactamases under antibiotic pressure , we conducted a selection experiment by exposing Burkholderia thailandensis [11] to ceftazidime ( 3–5 µg/ml ) . The resistant colonies were screened for isolates with variants of the penA gene that acquired a duplication within the coding region ( Figure 1A ) . PenA is a class A β-lactamase ( BTH_II1450 from B . thailandensis strain E264 ) that confers resistance to amoxicillin . We observed that the relative frequency of occurrence of the TR mutations compared to substitution mutations was about 1 to 50 . The frequency for substitution mutations was previously estimated to be 10−8 to 10−7 [12] . PenA has been used to explore evolutionary paths by various mutations to substrate spectrum extension [12] , [13] . PenA from B . thailandensis is highly conserved in pathogenic Burkholderia species including Burkholderia pseudomallei , Burkholderia mallei , and Burkholderia cenocepacia [14]–[16] . The antibiotic regimen used to treat infections by these Burkholderia pathogens generally includes ceftazidime [17] . The de novo TRs in penA from the ceftazidime-resistant isolates , which we named TR1 through TR11 , consisted of two repeat units ( the template and the duplicate ) that involve at least part of the region encoding the omega loop ( Figure 1A ) . Alterations of the omega loop caused by specific mutations have been implicated in substrate spectrum extension [18] , [19] . In a study in which a variable five amino acid cassette was randomly introduced into TEM-1 , all insertions that conferred enhanced resistance to ceftazidime were consistently found at various positions within the omega loop [20] . The lengths of the repeat units of these TRs varied widely ranging from 9 bp ( TR2 and TR3 , coding for 3 amino acids ) to 60 bp ( TR11 , coding for 20 amino acids ) . Except for TR2 , all TRs had a discrepancy between the repeat units and the reading frame ( codons ) in the gene , and this often resulted in a different amino acid in the first position of the repeated peptide ( Figure 1A ) . We discovered that three of the TRs ( those numbered 1 to 3 ) were associated with a pair of direct repeats ( Figure 1A; for nucleotide sequences of TRs and a comparison before and after duplication , see Figure S1 ) . Although the lengths varied from 3 to 8 , all had a common pattern in that the direct repeats were perfectly-matching to each other and the DNA template for duplication starts with the upstream repeat and ends just before the downstream repeat ( Figure 1A , Figure S1 ) . It is reasonable to assume that DNA duplication specifically defined by these direct repeats is mediated by the replication slippage mechanism established in various genomes [2] . We found that the other group of TRs ( those numbered 4 to 11 ) also had direct repeats but these repeats were not bordering the template as in TRs 1 to 3 , obscuring their role in DNA duplication ( Figure 1A ) . In addition to the direct repeats , these TRs had a different duo of sequences ( Figure 1A , Figure S1 ) . These sequences had an interesting pattern—each pair had a high degree of seeming complementarity ( Figure 1A ) . However , these sequences were on the same strand , and so we named them SCSs ( same-strand complementary sequences ) . Whereas five TRs ( 5 , 6 , 7 , 8 , and 10 ) had perfectly matching SCSs , others had the ones with mismatches ( Figure 1A ) . Besides the mismatches , SCSs also exhibited variations in their positions relative to the template for TRs ( Figure 1A ) . There undoubtedly would be selective pressure for a template during DNA duplication . The minimum requirement would be to maintain the reading frame in the gene with a multiple of 3 nucleotides in the repeat unit . Then the next selection requirement would be steric modification of the active site to accommodate new drugs without inducing serious protein structural instability . We found SCSs around TRs in the two clinical isolates obtained to date: in the SHV-16 gene from K . pneumoniae [5] and in the ampC gene , coding for a class C β-lactamase , from E . cloacae GC1 [7] ( Figure 1B ) . The SCSs in these genes were perfectly matching , in contrast to those with occasional mismatches in penA in B . thailandensis ( Figure 1A ) . No direct repeats were found around the DNA template in these genes , suggesting that the SCSs in TRs 4 to 11 may be functionally sufficient in mediating DNA duplication . These β-lactamase genes from distinct sources comprise the evidence supporting the identity of SCSs and that DNA duplication mediated by SCSs is a common mechanism for the substrate spectrum extension across β-lactamase types and bacterial groups . To test if SCSs mediate TR formation , we generated two point mutations in the wild-type penA in the region corresponding to the 3′-ends of the DNA templates for TR4 and TR5 , in the way not affecting the encoded amino acids ( see Materials and Methods , Figure 2A ) . This region also overlaps with the region at the 5′-ends of the templates for TRs 7 to 11 ( Figure 2A ) . Therefore , the substitution mutations can alter the seven pairs of SCSs ( Figure 2A ) . Against the selection with ceftazidime , the penA gene with the disrupted SCSs exhibited a similar TR pattern to that of the wild type in the region not containing the mutations ( that for TRs 1 , 2 , 3 , and 6 ) ( Figure 2B; for principal components analysis ( PCA ) plots [21] comparing the wild type and the mutant , see Figure S2A ) . However , a significantly altered pattern was observed around the point mutations in that six of the seven TRs were not formed in the mutated gene ( Figure 2B , Figure S2A ) . The only one that was still formed in the mutant was TR10 ( Figure 2B ) . However , this formation of TR10 in the mutated gene may have been enabled by a new pair of SCSs generated around the DNA template , complementing the disrupted original SCS pairs ( Figure 2C ) . In addition , eight new TRs were formed in the mutated gene ( Figure 2C; for a comparison between before and after duplication , see Figure S2B ) . Two of these , M-TR1 and M-TR2 , occurred via direct repeats made with the new T nucleotides , demonstrating the mechanistic role of direct repeats specifically bordering the DNA template as in TRs 1 to 3 ( Figure 1A ) . Likewise , M-TRs 3 to 8 apparently were formed by a pair of common SCSs made with the new T's along with a pair of neighboring SCSs , providing a basis for strong SCS interactions in the region ( Figure 2C ) . These M-TRs may demonstrate the functional role of SCSs in DNA duplication , and also that different regions can be used as templates during DNA duplication mediated by the same set of SCSs . In fact , TR7 and TR8 were also resulted from duplication of slightly different template areas mediated by the same set of SCSs ( Figure 1A ) . In the case of M-TR4 , however , a short pair of direct repeats with TR1-type positioning was also present with the possibility of the alternative mechanism ( Figure 2C ) . Together , the altered TR patterns in the SCS-altered gene strongly support the active role of SCSs as well as direct repeats with TR1-type positioning in DNA duplication in the β-lactamase gene . The minimum inhibitory concentration ( MIC ) for ceftazidime of the TR mutants were in the range of 7 . 3 ( TR4 ) to 24 ( TR5 ) ( Figure 3 ) . Significant correlation between the MICs and the size or the location of the duplication was observed ( Figure 3 ) . When the penA gene with TR10 or TR11 ( penA-TR10 or penA-TR11 , respectively , representing the TR-containing genes ) was disrupted ( see Materials and Methods ) , the strain lost resistance to ceftazidime ( MIC≤1 . 0 µg/ml ) , having an MIC comparable to the level of the wild-type strain with ΔpenA ( Figure 3 ) . When these strains were provided with an intact copy of the penA-TR10 or penA-TR11 in trans , ceftazidime resistance was restored ( Figure 3 ) , demonstrating that penA-TRs were the sole factors responsible for ceftazidime resistance . It is notable that resistance to ceftazidime can be dramatically increased through increased expression of the β-lactamase gene either when the gene is placed on a multicopy plasmid ( Figure 3 ) or by a point mutation in the promoter of the gene [12] , [22] . Substrate spectrum extension by mutations in β-lactamases is typically accompanied by a loss of the catalytic activity in the original substrates [8] . Accordingly , we observed decreased levels of MICs with one of the original antibiotic , amoxicillin ( Figure 3 ) . Three additional antibiotics ( cefotaxime , ceftriaxone , and cefepime ) that could be hydrolyzed by the wild-type enzyme [12] were tested with the bacteria harboring either of the two penA-TRs ( penA-TR10 and penA-TR11 ) ( Table 1 ) . These strains also exhibited decreased levels of resistance to these original substrates , as has been observed with many class A β-lactamase mutants with acquired activity against third-generation cephalosporins [23] . In addition , the hydrolytic activities against amoxicillin were effectively inhibited by a β-lactamase inhibitor clavulanic acid ( Table 1 ) . The original and the extended substrate spectrums of a β-lactamase define two phases of distinct catalytic activity [8] , [24] . The reversion back to the wild type of most mutations including point mutations is highly unlikely at typical mutation rates , especially when the associated fitness cost is not significant [25] . Accordingly , we failed to observe reversion in penA with the point mutations Cys69Tyr and Asp179Asn [12] under selection against amoxicillin at a concentration that was tolerated by the wild-type strain but not by the mutants ( 16 µg/ml , Figure 4 ) . In contrast , when the strains carrying penA-TR10 or penA-TR11 were challenged with the same antibiotic pressure , revertants were observed ( Figure 4A ) . The average reversion frequency of the strain carrying penA-TR11 with a longer repeat was 3 . 6×10−5 , which was more than ten-fold higher than that of the strain carrying penA-TR10 ( 9 . 3×10−7 , Figure 4B ) . The reversion capacity of the TR mutants of β-lactamase genes may be based on the intrinsic instability of TRs in genomes , as has been previously reported in various bacteria and eukaryotes including humans [1] , [3] , [26] . The instability in TRs in bacterial genomes is based on general cellular pathways including replication , recombination , and DNA repair systems [2] , [4] , [26] . Bacterial cellular pathways ( recombination in particular ) have been the focus of attention for alternative drug targets to prevent the generation and spread of antibiotic resistance [27] , [28] . The TR-based antibiotic resistance further lends support to such approaches in the battle against bacterial antibiotic resistance . We scanned the penA gene and the entire genome of B . thailandensis for direct repeats and SCSs , and found that the region encoding the omega loop in penA has unusually high contents of both sequences ( Figure 5 ) . It was particularly distinct with SCSs in that the omega loop and the immediate downstream region had the highest number for the sequences . These data suggest that there is high DNA duplication potential in the omega loop region , and support the significance of the region as the hotbed for the evolution of the β-lactamase . This omega loop region with a high potential of mutations makes the β-lactamase gene a so called “contingency gene” that facilitates the efficient exploration of phenotypic solutions to unpredictable ( host ) environment [3] . As direct repeats have been implicated in DNA duplication at the genomic level [2] , a high correlation is expected between the direct repeats content and DNA duplication activities in genomes . Accordingly , we observed a high Pearson correlation coefficient ( 0 . 9 ) between the number of direct repeats and that of TRs among the Burkholderia spp . —the hosts for the penA gene ( Figure 6A ) . We also measured a high correlation ( Pearson correlation coefficient of 0 . 99 ) between the number of SCSs and that of TRs across genomes ( Figure 6A ) . This suggests that SCSs are also associated with DNA duplication activities in Burkholderia at the genomic level , and that DNA duplication mediated by both direct repeats and SCSs comprise a general mechanism in Burkholderia . Then we expanded the analysis to include diverse bacterial genomes . All completed bacterial genomes were downloaded from the National Center for Biotechnology Information ( NCBI ) and 1 , 387 genomes ( those having the largest genome in a species ) were selected as representatives for each species . These genomes were grouped in genera , and the ones with at least five members ( species ) , which were a total of 59 genera , were subjected to the analysis . In the analysis comparing the contents of SCSs and TRs among the members within each genus , we found 42 genera with at least 0 . 7 of the Pearson correlation coefficient , including 25 genera with a highly significant correlation ( >0 . 9 ) ( Figure S3 ) . However , there were also the genera with poor correlations ( Figure 6B; Figure S3 ) . Genera with similar correlation coefficients , especially distinctive with the low ones , showed a tendency to form clusters in the phylogenetic tree , and the GC contents in the genomes had no significant correlation with such patterns ( Figure 6C ) . This phylogenetic profile could be interpreted as partial or total loss of the SCSs-mediated DNA duplication mechanism in some bacterial groups during the course of the evolution . An analysis with direct repeats also revealed the distribution of the high correlation with TR levels among the most bacterial genera , with lower correlations observed in some groups ( Figure S3 ) . Together , the direct repeats-TRs and SCSs-TRs correlation patterns in the bacterial kingdom suggest that DNA duplication mediated by direct repeats and SCSs comprises a fundamental mechanism shaping bacterial genome evolution . In this study we characterized the formation of DNA duplication in the coding region of a β-lactamase gene . The de novo TRs , coupled with reversion , has been adapted to mediate reversible switching between the two states of the β-lactamase substrate spectrums for bacterial survival against dynamic antibiotic challenges ( Figure 7A ) . A characteristic attribute of this novel toggling mechanism is its function on enzyme activities . This is in contrast with the toggle switch constructed with genes arranged in a mutually inhibitory network , by which transcription of the genes is alternated [29] . As a practical sense , the TR-based toggles may have implications for biotechnology for the use to reversibly affect the activity of a target protein . The direct repeats at the specific positions relative to the DNA template—as those in TRs 1 to 3 and M-TRs 1 and 2—fit perfectly to the known mechanism , DNA replication slippage [2] ( Figure 7B ) . By contrast , the role of the direct repeats without such positioning—as those in TRs 4 to 11—is not clear . Unlike direct or inverted repeats , SCSs would not interact with each other through usual Watson-Crick base pairing . This suggests that a novel form of DNA conformation would be involved during this DNA duplication , possibly through reverse Watson-Crick or reverse Hoogsteen base pairing [30]–[33] ( Figure 7B ) . Non-canonical base interactions have been found more prevalent in DNA than previously predicted , and in some DNA sequences Hoogsteen base pairs are in thermal equilibrium with Watson-Crick base pairs [34] . If such SCSs-based DNA structures are formed ( Figure 7B ) , facilitated by unknown factors and/or by certain local topological strain , they may cause structural instability in the region , causing mistakes in the DNA replication , repair , or recombination machineries , predisposing to DNA duplication [26] , [30] . Although the underlying mechanism is not understood , the biological consequences resulting from the SCSs-mediated DNA duplication are significant . In this study , we showed that more TRs were mediated by SCSs than by direct repeats in the β-lactamase gene , penA , during selection ( Figure 1A , Figure 2ABC ) . That SCS-TRs may be common across different types of β-lactamases ( the classes A and C ) and bacterial groups ( Figure 1B ) , and that SCSs may be involved in more fundamental processes shaping the bacterial evolution ( Figure 6 ) signifies the importance of SCSs . It is intriguing to note that sequences corresponding to SCSs were also found over-represented in the flanking regions of insertion and deletion mutations in various eukaryote genomes [35] , and association of such sequences with human genetic diseases was suggested [36] . The potential pivotal role of SCSs in the evolution of both prokaryotic and eukaryotic genomes and in human diseases further assure the need for the active investigations on SCSs .
All Escherichia coli strains were grown in Luria Bertani ( LB ) media , and all B . thailandensis strains were grown in LB or AB minimal media containing 0 . 25% glucose ( ABG ) [37] at 37°C . The concentrations of antibiotics used for E . coli were as follows: tetracycline , 10 µg/ml; kanamycin , 50 µg/ml; and ampicillin , 100 µg/ml . For B . thailandensis , the concentrations of tetracycline and kanamycin used were 50 µg/ml and 250 µg/ml , respectively . To map the mutations in penA ( BTH_II1450 in the strain E264 genome ) , we began by sequencing the gene . Genomic DNA from each ceftazidime-resistant mutant was purified using a Wizard Genomic DNA Purification Kit ( Promega , Madison , WI , USA ) and was used as the template for the PCR amplification of penA and its short flanking regions to yield a 1 , 386-bp amplicon ( 271 bp upstream from the start codon of the gene to 230 bp downstream of the stop codon ) . PCR reactions were conducted in a 50-µl reaction mixture containing 2 . 5 U of HotStar HiFidelity polymerase ( Qiagen , Hilden , Germany ) , 50 pmol of the primers penA-F ( 5′-CGTCAATCCGATGCAGTACC-3′ ) and penA-R ( 5′-GCCGTTATCGCACCTTTATC-3′ ) , 100 ng of template DNA , 10 µl of 5× Q Solution , and 10 µl of 5× HotStar HiFidelity buffer . The reaction protocol consisted of an initial enzyme-activating step ( 95°C for 5 min ) , an amplification step ( 35 cycles at 94°C for 15 sec , 61°C for 1 min , and 72°C for 1 . 4 min ) , and a final extension step ( 72°C for 10 min ) . Gel-purified 1 . 4 kb PCR products were sequenced using a 3730XL DNA analyzer ( Applied Biosystems , Foster City , CA , USA ) in both directions using the primers penA-F and penA-R . The MIC values were measured using the E-test [38] , following the manufacturer's instructions ( AB Biodisk , Solna , Sweden ) . Briefly , each strain was grown on Müller-Hinton agar plates at 37°C for 2 days . For the strains harboring pRK415K-derived plasmids , Müller-Hinton agar supplemented with kanamycin ( 250 µg/ml ) was used . Single colonies from the plates were suspended in 2 ml of 0 . 85% NaCl until the turbidity reached the 0 . 5 McFarland standard . Using sterile cotton swabs , the cell suspensions were spread on Müller-Hinton agar plates , the E-test strips were placed , and the plates were incubated at 37°C for 16–18 h . Using the E-test strips , the lowest concentration where no visible growths was observed was recorded as the MIC . The average values were calculated from triplicate experiments . The agar dilution method was conducted as described by Wiegand et al . [39] . Briefly , a single colony of each strain grown on Müller-Hinton agar was used to inoculate 2 ml of Müller-Hinton broth , and the culture was incubated overnight with shaking ( 250 rpm ) at 37°C . The overnight cultures were serially diluted with fresh Müller-Hinton broth and dispensed into the wells of a 96-well microtiter plate . Using a multi-channel micropipette , 1 µl of the diluted bacterial suspensions was placed on each Müller-Hinton agar plate containing ceftazidime at various concentrations and incubated at 37°C for 16 h . The lowest concentration of antibiotic such that there was no visible bacterial growth was observed in the spot containing approximately 104 CFU ( colony forming unit ) , which was recorded as the MIC . The number of CFUs in serially diluted bacterial suspensions was determined by spreading 100 µl of the appropriately diluted bacterial suspensions on LB agar plates , followed by incubating the plates for 24 h at 37°C , and counting the colonies derived from viable cells . The average values were calculated from triplicate experiments . The wild-type and mutated penA genes were disrupted as follows . First , the wild-type penA was PCR-amplified using the primer pairs penA-KF ( 5′-ATATATGGTACCCGTCAATCCGATGCAGTACC-3′ ) and penA-KR ( 5′-ATATATGGTACCGCCGTTATCGCACCTTTATC-3′ ) that contained a KpnI recognition site ( underlined ) at the end . The 1 . 4-kb PCR product was then digested with KpnI and ligated into pUC19 , which was digested using the same enzyme . The resulting plasmid was double-digested with XhoI and PflFI to remove an 195-bp internal region ( position 124 bp to 319 bp ) in penA , blunt ended with T4 DNA polymerase ( NEB , Ipswich , MA , USA ) , and inserted with a tetracycline-resistance ( tetr ) cassette by ligation . The tetr cassette was previously amplified from pRK415K [40] using HotStar HiFidelity polymerase ( Qiagen , Hilden , Germany ) and primers tetR-F ( 5′-ATATATCTCGAGGTGAGGCTTGGACGCTAGG-3′ ) and tetR-R ( 5′-ATATTTCTCGAGCTTGGATCAGACGCTGAGTG-3′ ) that contained an XhoI recognition site ( underlined ) , XhoI-digested , and blunt-ended with T4 DNA polymerase . The construct was transferred into B . thailandensis strains employing a modified method of natural transformation [41] . Briefly , 3 ml of a defined medium ( DM ) that was prepared as described by Thongdee et al . [41] was inoculated with a single colony freshly grown on LB agar and then incubated overnight with shaking at 250 rpm at 37°C . The 200-µl overnight culture was diluted with 10 ml of fresh preheated DM , and the culture was grown with shaking at 250 rpm at 37°C to an OD600 of approximately 0 . 5 . The culture was then concentrated 20-fold in 500 µl of fresh DM and 50 µl aliquots of the concentrated cells were mixed with 0 . 5 µg of plasmid DNA . The mixture was incubated for 30 min on ice , and 2 ml of DM preheated to 37°C was added before the mixture was incubated overnight at 250 rpm at 37°C . After washing the culture with 1 ml of fresh DM and resuspending the pellet in 250 µl of fresh DM , 100 µl of the cell suspension was plated on ABG medium containing 50 µg/ml tetracycline and incubated at 37°C for 48 hrs to select for tetr cassette-containing constructs . An obtained penA null mutant was verified by PCR using the primer pair of penA_LF ( 5′-AACAGATCGCCGAGATGG-3′ ) and penA_LR ( 5′-GCGAACGTTGCCCGATAC-3′ ) , which hybridizes to the genomic regions outside the DNA sequence that is used for mutant construction . KpnI-treated PCR products of mutated penA , which was amplified using the primers penA-KF and penA-KR as described above , were ligated with KpnI-treated pRK415K . This plasmid was transferred into the E . coli strain S17-1 [42] using a conventional transformation method [43] . The transformed S17-1 strain was conjugated with a B . thailandensis penA-null mutant on ABG agar plates containing 250 µg/ml kanamycin , and the plates were incubated at 37°C for 2 days to select for transconjugants . A successful conjugation was confirmed by purifying the plasmid from the transconjugants and examining the characteristic restriction patterns of the plasmid . To estimate the reversion frequency of the mutations in penA , we compared the colony-forming units of the serially diluted strains on LB plates without ( for total cell counts ) and with amoxicillin at a final concentration of 16 µg/ml ( for revertant counts ) . The alleles of penA in the revertants were sequenced for verification . Each of the serially diluted strains was prepared as follows . First , each of the strains was grown on an LB agar plate for two days , and a colony was then suspended in LB at a concentration of approximately 2×105/µl . This cell suspension was then serially diluted in LB . Five microliters of the cell suspension and 5 µl of each dilution were spotted on LB plates , and the plates were incubated for approximately 24 to 30 hours until colonies formed . The strain with penA gene with two substitution mutations in a common SCS element ( CACGGCG to CACTGCT ) ( Figure 2A ) was constructed as follows . The DNA fragment of 1 , 445 bp , which spans the whole penA-TR11 gene and the extra flanking sequences was constructed by ligating two separately prepared fragments . The first half of the gene , which spans from the position 271 bp upstream to the position 527 bp downstream from the start codon , was obtained using PCR conducted in a 50 µl reaction mixture containing 1 U of KOD FX NEO polymerase ( TOYOBO , Japan ) , 15 pmol each of the primers penA-KF ( 5′-ATATATGGTACCCGTCAATCCGATGCAGTACC-3′ ) , containing a KpnI recognition site ( underlined ) , and penA-R-527 ( 5′-GTGTTCAGTTCGGTCTCGCG-3′ ) , 100 ng of the genomic DNA from the strain containing penA-TR11 , 25 µl of 2× PCR buffer for KOD FX Neo and 10 µl of 2 mM dNTPs . The reaction consisted of the following three steps: initial pre-denaturation step ( 94°C for 2 min ) , amplification step ( 35 cycles of 98°C for 10 sec , 64°C for 1 min , and 68°C for 48 sec ) , and the final extension step ( 72°C for 10 min ) . The second half of the gene spans from the position 528 bp downstream from the start codon to the position 229 bp downstream from the stop codon . This fragment was amplified as described above except for using a different pair of primers: penA-T2-528 ( 5′-TGCTATTCCCGGCGACGAGC-3′ ) , in which two point mutations ( underlined ) were introduced and penA-ER ( 5′-ATATATGAATTCGCCGTTATCGCACCTTTATCGC-3′ ) , which contains an EcoRI recognition site ( underlined ) . The resulting two fragments were phosphorylated with polynucleotide kinase ( NEB , Ipswich , MA , USA ) and were digested with KpnI and EcoRI , respectively , and the fragments were cloned into KpnI and EcoRI digested pUC19 . The ligated penA-TR11 gene with introduced substituted mutations was cut out of the pUC19 construct by digesting the construct with KpnI and EcoRI , and the gene fragment was blunt-ended with T4 DNA polymerase ( NEB , Ipswich , MA , USA ) . Then the gene fragment was cloned into a suicide vector , pSR47S [44] . The resulting pSR47S construct was introduced into Burkholderia thailandensis E264 using triparental mating . Briefly , the recipient strain Burkholderia thailandensis E264 was grown in LB overnight , and the donor strain E . coli DH5αλpir containing pSR47S construct and the helper strain E . coli containing pRK600 were grown overnight in LB supplemented with kanamycin at 50 µg/ml and chloramphenicol at 20 µg/ml , respectively . The overnight cultures of the three strains were washed with PBS and were mixed to the final volume of 100 µl , and the mixture was spread on an LB agar plate and was incubated at 37°C for 3 hours . Then the cells were resuspended in 2 ml of PBS and were spread on an LB agar plate supplemented with ceftazidime at 5 µg/ml and 5% sucrose and were incubated for 48 hours at 37°C . The obtained strain containing the allele-exchanged penA gene with substitution mutations was plated on an LB agar plate supplemented with amoxicillin at 16 µg/ml , to select for a revertant , which has the penA gene with two substitution mutations . Allele exchange and reversion were verified by carrying out PCR with primers penA_OF ( 5′-AACAGATCGCCGAGATGG-3′ ) and penA_OR ( 5′-GCGAACGTTGCCCGATAC-3′ ) , which anneal to the regions outside the cloned region , and by sequencing the PCR product . Bacterial genomes were downloaded from the National Center for Biotechnology Information ( NCBI ) . For simpler computations , SCSs were confined to those with perfect complementarity with a minimum unit size of 6 bps and a distance between the units of 6 to 100 bps . Direct repeats were confined to those with a minimum unit size of 5 bps and have a distance between the repeats of 6 to 100 bps . Customized perl scripts were used to identify the SCSs and direct repeats in genomes . For TRs , only those with the unit length equal to or larger than 6 bases ( capacity for a peptide with two amino acids ) were counted . TRs were identified using the software tool Tandem Repeat Finder [45] . To construct the phylogenetic tree with selected bacterial genera , 16S rDNA sequences of the selected genera were aligned using PyNAST [46] . Gaps were removed using a Qiime script ( http://www . qiime . com ) , and then the tree was built using clustalX2 ( http://www . clustal . org/clustal2/ ) . | β-lactamases can adapt to new antibiotics by mutations in their genes . The original and the extended substrate spectrums of β-lactamases define two phases of catalytic activity , and the conversion by point mutations is unidirectional from the initial to the new spectrum . We describe duplication mutations that enable reversible switching between the substrate spectrums , increasing the adaptability of the bacterium . We provide evidence supporting that two distinct groups of short sequences mediated the formation of DNA duplications in β-lactamases: direct repeats and novel elements that we named , SCSs ( same-strand complementary sequences ) . Our study suggests that DNA duplication processes mediated by both direct repeats and SCSs are not just limited to the β-lactamase genes but comprise a fundamental mechanism in bacterial genome evolution . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"bacteriology",
"antimicrobials",
"organismal",
"evolution",
"microbial",
"mutation",
"medicine",
"and",
"health",
"sciences",
"burkholderia",
"infection",
"microbiology",
"bacterial",
"diseases",
"antibiotics",
"microbial",
"evolution",
"molecular",
"genetics",
"bacterial",
... | 2014 | The Tandem Repeats Enabling Reversible Switching between the Two Phases of β-Lactamase Substrate Spectrum |
Dynamics and functions of G-protein coupled receptors ( GPCRs ) are accurately regulated by the type of ligands that bind to the orthosteric or allosteric binding sites . To glean the structural and dynamical origin of ligand-dependent modulation of GPCR activity , we performed total ~ 5 μsec molecular dynamics simulations of A2A adenosine receptor ( A2AAR ) in its apo , antagonist-bound , and agonist-bound forms in an explicit water and membrane environment , and examined the corresponding dynamics and correlation between the 10 key structural motifs that serve as the allosteric hotspots in intramolecular signaling network . We dubbed these 10 structural motifs “binary switches” as they display molecular interactions that switch between two distinct states . By projecting the receptor dynamics on these binary switches that yield 210 microstates , we show that ( i ) the receptors in apo , antagonist-bound , and agonist-bound states explore vastly different conformational space; ( ii ) among the three receptor states the apo state explores the broadest range of microstates; ( iii ) in the presence of the agonist , the active conformation is maintained through coherent couplings among the binary switches; and ( iv ) to be most specific , our analysis shows that W246 , located deep inside the binding cleft , can serve as both an agonist sensor and actuator of ensuing intramolecular signaling for the receptor activation . Finally , our analysis of multiple trajectories generated by inserting an agonist to the apo state underscores that the transition of the receptor from inactive to active form requires the disruption of ionic-lock in the DRY motif .
G-protein coupled receptors ( GPCRs ) are one of the most versatile membrane proteins that mediate cellular responses to a myriad of extracellular signals associated with our perception , cardiovasicular , and immune functions [1] . GPCRs relay extracellular signals to the cytoplasmic domain . For a given extracellular signal , there is a corresponding GPCR subtype that processes the incoming signal to the cellular downstream [2] . Consisting of seven transmembrane ( TM ) helices , each of which is connected to the next TM helix by either an intracellular loop ( ICL ) or an extracellular loop ( ECL ) , the interior of GPCR forms an interhelical residue-to-residue interaction network that can transmit the signal specific to the ligand and/or receptor subtype . Binding of an agonist to the orthosteric site leads to conformational rearrangement of TM helices , transforming the inactive conformation to the active one , which in turn enables accommodation of G-proteins and intracellular signal transductions [1 , 3 , 4] . It is widely appreciated that together with highly conserved residues in the TM region , the activation of the receptors belonging to the class A GPCR family is regulated by a set of fingerprint residues called “microswitches” [1 , 5] , which commonly include three structural motifs: DRY , CWxP , and NPxxY motifs ( “x” denotes any amino acid residue ) . Upon binding of an agonist , the microswitch residues change the orientation of their side chain , which transforms the global configuration of TM helices into the active form and helps the intracellular domain accommodate G-proteins [5] . In contast , binding of an inverse-agonist suppresses the GPCR function below the basal level , to which the apo or antagonist-bound forms of GPCRs is likely to be tuned [1] . The allostery , a long-range communication between two remote sites [6–9] , is one of the key determinants for functions in many biomolecules . And it is of particlular interest for GPCRs because most of dynamic processes associated with GPCR activation and suppression involve signaling between extracellular and intracellular domains , established across the TM domain . Although the term “allosteric modulation” is often strictly distinguished from the orthosteric signaling in GPCR community , both can be considered allosteric signaling in that their signalings are both physically long-ranged . While high-resolution crystal structures in the agonist-bound or antagonist-bound states provide unprecedented view of ligand-dependent modulation of GPCR conformation , minor conformational difference between distinct receptor states make it difficult to glean the microscopic origin of intramolecular signaling and modulation of GPCR activity . Unlike molecular machines [10–13] and enzymes that exhibit large conformational changes [14 , 15] , to which one can apply a method such as normal mode analysis and its variation [16 , 17] , it is not straightforward to study the allosteric dynamics of biomolecules like GPCRs[18–21] whose structural changes between the inactive and active forms are relatively small ( RMSD between inactive and active forms are only 1 . 78 Å for A2A adenosine receptor; and 2 . 96 Å for β2 adrenergic receptor whose active state structure is crystalized with Gs-protein ) . It is noteworthy that there is a growing realization that a certain class of proteins can display allosteric responses by modulating conformational fluctuations , but with little conformational change [22–25] . Global conformational changes themselves are not the sole physical origin of protein allostery . Thus , to gain a better understanding of allostery or long-range intramolecular signaling of GPCR , it is imperative to probe the dynamical features of key structural components and their correlations . Although experimental methods , such as fluorescence resonance energy transfer ( FRET ) , and site-directed spin labeling ( SDSL ) , nuclear magnetic resonance ( NMR ) [26–30] , are useful to monitor the conformational changes of specific key residues , it is generally difficult to simultaneously probe the multiple sites of molecule and elucidate dynamic origin leading to allosteric molecular responses . Much effort has recently been made to study the mechanism of GPCR activation via molecular dynamics ( MD ) simulations [31–36] by exploring the dynamic characteristics of GPCR conformations; however , systematic studies of cross-correlations among key structural motifs including microswtiches have not been carried out . From sequence analysis , structural and biochemical studies using mutagenesis , there is a general consensus for the class A GPCRs that 18 hotspot residues , called microswitches , are responsible for the activation mechanism [1 , 3] . Recently we have shown that when GPCR structures are represented by a network of inter-residue contacts , many of those microswitches [1 , 5] retain high betweenness centrality values [21] . According to the network theory , vertices with high betweenness centralities in a given network are the sites that mediate flow of signal over the network [37 , 38] . A removal or alteration of such vertices from the network , which can be realized in the form of deletion [39] or more practically mutation into glycine [21] , could impair the network communication and be deleterious to intramolecular signaling . Among the 18 microswitches , our network analysis in Ref . [21] could identify 11 of them in A2A adenosine receptor ( A2AAR ) , and also showed that the majority of intra-molecular signaling pathways connecting the highly correlated residues between extra and intra-cellular domains pass through the 11 microswitches . However , our previous study provided only a static view of intramolecular allosteric wiring . Thus , to better understand the intramolecular signaling of A2AAR , it would be of great interest to directly probe the dynamics associated with these microswitches and their cross-correlations . In this study , we performed each of 1 μsec all-atom molecular dynamics simulation of A2AAR in apo , antagonist-bound ( ZM-241385 ) , [2 , 40] and agonist-bound ( UK-432097 ) [41] forms ( see Methods ) ; and monitored the local dynamics of the microswitches . Based on the dynamics around the 11 microswitches , each of which displays molecular interactions that switches between two distinct states , we defined 10 ON/OFF binary switches as the microscopic components to describe the interhelical dynamics ( note that the two microswitch residues of ionic-lock compose one on/off switch ) . Describing the dynamics of the receptor by employing 10 microscopic components amounts to projecting the entire dynamics into 10 local variables and considering 210 microstates . By projecting the receptor dynamics on 210 microstates , we show how the dynamics of each ligand-dependent macrostate differs from each other . Our simulations and analysis show that compared to that of the apo and antagonist-bound forms , the 10 binary switches of agonist-bound form retain greater cross-correlation and coherence , which is consistent with the notion that GPCR in agonist-bound form has a functional fidelity to transmitting its intra-molecular signals across the TM domain [21] . Detailed knowledge on the local dynamics and their dynamic correlation elucidated in this study for A2A adenosine receptors could be useful for the discovery of effective drugs .
Among the 18 microswitches of A2AAR , our previous study calculating the betweenness centrality of residue interaction network suggested that 11 of them are the putatively important spots for the intramolecular signal transmission [21] , which include N241 . 50 , D522 . 50 , D1013 . 49 , R1023 . 50 , W1294 . 50 , Y1975 . 58 , E2286 . 30 , W2466 . 48 , N2847 . 49 , P2857 . 50 , Y2887 . 53 where the superscripts refer to the Ballesteros Weinstein numbering system [42] . Examining the local dynamics around the eleven allosteric hotspot residues [21] at different receptor states , we have defined 10 binary ( on/off ) switches . Below we first explain how we have defined each switch to be binary , which maximally separates the dynamic features of antagonist-bound state from those of agonist-bound state . Ten binary switches defined from eleven hotspot residues for intra-molecular signaling . Microswitches at the interfaces between TM1 , TM2 , and TM7 ( Switches 1 & 2 ) : In the class A GPCR family N24 and D52 are the most conserved microswitch residues in TM1 and TM2 , respectively , with high betweenness centrality values [21] . Comparison of the H-bond network between three different receptor states reveals notable differences in the receptor configuration around them ( Fig . 1 and Fig . S1 ) . First , in the apo form , H-bonds of N24 and D52 with S2817 . 46 contribute to the stable conformation of TM helices . Binding of antagonist further stabilizes this conformation by incorporating an additional H-bond between N24 and D52 . By contrast , binding of an agonist leads to the disruption of the H-bond between D52 and S281 ( HBD52−S281 ) and forms a new H-bond between S281 and N2807 . 45 , which is responsible for helix bending in TM7 ( see below ) . Ligand-dependent switching dynamics of the H-bond of S281 from one side to the other is reminiscent of the salt-bridge switching , which is often observed at the interface between subunits in multisubunit proteins [10 , 43] . Hence , we designate HBN24-D52 and HBD52-S281 as the switch 1 ( 𝓢 1 ) and the switch 2 ( 𝓢 2 ) , respectively . DRY motif and ionic-lock ( Switches 3 & 4 ) : The orientation of the TM5-ICL3-TM6 relative to the ICL2 is the hallmark of GPCR activation as the 10o tilt of TM5-ICL3-TM6 enables the G-protein accommodation . Among others , the “ionic-lock” , the salt-bridge between R102 and E228 , which keeps the TM3 and TM6 in close proximity , is considered the key structural motif that directly regulates the orientation of TM6 helix [44] . Our simulations find that , in the antagonist-bound state , the ionic-lock is intact ( dR102Cζ-E228Cδ ≈ 4 Å ) ( Fig . 2a and the top panel of Fig . 2b ) ; while it is rarely formed in the agonist-bound state ( dR102-E228 ≈ 10 Å ) . In the apo state , the ionic-lock maintains the identical distance with that of the antagonist-bound state , but occasionally breaks ( dR102Cζ-E228Cδ ≈ 7 Å ) and reforms ( dR102Cζ-E228Cδ ≈ 4 Å ) ( Fig . 2 and Fig . S2 ) . The ionic-lock and the inter-residue distance ( dL110-A221 ) between L110Cα in ICL2 and A221Cα in ICL3 were simultaneously probed in Fig . 2b . It highlights the correlation between the ionic-lock and the orientation of TM5-ICL3-TM6 relative to TM3 . Dependence of dL110-A221 on the receptor state is clearly observed; dL110-A221 ≈ 22 Å ( agonist-bound ) , 7 Å ( antagonist-bound ) , and 14 Å ( apo ) , which quantifies the position of TM6 helix relative to TM3 helix . For the apo form , the distance between ICL2 and ICL3 varies concomitantly with the state of ionic-lock ( see the time traces after 500 nsec of Fig . 2b ) . The scatter plot of dR102-E228 and dL110-A221 ( right panel of Fig . 2b ) also shows a clear correlation between the dynamics of ionic-lock and the orientation of TM5-ICL3-TM6 domain . The status of ionic-lock affects the H-bond network around the DRY motif ( Fig . 2c and Fig . S3 ) . In the agonist-bound form , E228 forms the H-bond with R107 instead of R102 . Disruption of ionic-lock leads to release of TM6 from TM3 , but as a counterbalance a new H-bond ( HBT41-D101 ) binds TM3 tightly with TM2 . As HBT41-D101 and the ionic-lock in the DRY motif can be used to discern the ligand-dependent macrostate of GPCR , we define HBT41-D101 and the ionic-lock as 𝓢 3 and 𝓢 4 , respectively . Rotameric microswitches ( Switches 5 , 6 , 7 , 8 , & 10 ) : Ligand-dependent microstates of the five residues W129 , Y197 , W246 , N284 , and Y288 are best characterized by the rotameric angles of their side chains ( Fig . S4 ) , which allow us to define these residues as 𝓢 5 , 𝓢 6 , 𝓢 7 , 𝓢 8 , and 𝓢 1 , respectively: ( i ) In the agonist-bound state , W129 ( 𝓢 5 ) shows asymmetric bimodal distribution of dihedral angle P ( χ2 ) that has a major peak at ≈ −20o and a minor peak at ≈ 90o . In the apo and antagonist-bound forms , P ( χ2 ) is unimodal with a peak at 90o ( Fig . 3a and Fig . S5a ) ; ( ii ) The dihedral angles of Y197 ( 𝓢 6 ) also show distinct distribution in the three receptor states ( Fig . 3b and Fig . S5b ) ; ( iii ) W246 ( 𝓢 7 ) , known as a central switch located at the bottom of the orthosteric binding cleft [5] , exhibits receptor state-dependent rotamer angle dynamics between ≈ +90o and ≈ −90o ( Fig . 3c , Fig . S5c , and Supporting Movie M1 ) ; ( iv ) N284 ( 𝓢 8 ) and Y288 ( 𝓢 10 ) in the NPxxY motif show receptor-state dependent dihedral angle distribution ( Fig . 3d and Fig . S5d ) . Meanwhile , the dihedral angle distributions of P285 , which also belongs to NPxxY motif , are little affected by the receptor state . Thus , we use the kink angle of P285 instead of its dihedral angle for defining the switch ( see below ) . Helix bending induced by proline kink ( Switch 9 ) : The geometry of proline residue constrains the backbone conformation and cause a sharp kink in alpha helix structures . Calculation of the helix bending angle [45] shows that in TM7 helix , the greatest kink is formed around P285 , in particular , in the agonist-bound form ( Fig . 3e ) . Comparison between structures for different receptor states suggests that the H-bonding between N2807 . 45 and S2817 . 46 , in the agonist-bound state , contributes to the helix bending in TM7 . The scatter plot in Fig . 3e indicates that there is a direct correlation between TM7-helix bending ( θTM7 ) and HBN280-S281 . Projection of GPCR dynamics onto 10-binary switches . The dynamic features of the 11 microswitch residues that display two-state switch-like molecular interactions allow us to define 10 binary switches ( Fig . 4 ) . We choose the value of the i-th switch ( si ) as 1 or 0 ( ON or OFF ) in reference to the switch state in the agonist-bound form , so that the “similarity” of each switch can be assessed in reference to the agonist-bound form . For example , the HBD52-S281 , corresponding to 𝓢 2 , is disrupted in the agonist-bound form . In this case we set s2 = 1 for disrupted H-bond and s2 = 0 for intact H-bond . Also , for 𝓢 7 whose configuration is best described using rotameric angle , we consider 𝓢 7 to be in the ON state if the dihedral angle ( Cα-Cβ-Cγ-Cδ1 ) of W246 ( see Fig . S4 ) is equal or less than −50o; otherwise , it is in the OFF state ( s7 = 0 ) . Our 10 switch representation resembles the strategy in studying protein folding problem using “correctness” of the configuration of each residue with respect to the native state [46] . Representing GPCR conformation in terms of the 10 binary switches amounts to “choosing” multiple progress coordinates ( or multi-dimensional order parameters ) to probe the allosteric dynamics of GPCR from the inactive to active state . The assumption that 10 binary switch can faithfully represent the dynamics of GPCRs leads to in total , 210 possible microstates; each microstate is expressed using binary number from 0000000000 ( 2 ) to 1111111111 ( 2 ) with each digit denoting the switch number from 1 to 10 . These binary numbers can also be expressed with a decimal number from 0 and 1023 ( Fig . S6 ) . The time traces projected on the 10 binary switches and 210 microstates are shown in Fig . 5a and Fig . 5c , respectively . The average value of each switch , calculated in different receptor state ξ as 0 ≤ ⟨sξ , i⟩ ≤ 1 ( Fig . 5b ) , where ⟨sξ , i⟩ with i denoting the switch index is the value of switch averaged over the simulation time , indicates that on average switches are ON in the agonist-bound form , OFF in the antagonist-bound form , and they lie in between in the apo form . The difference among the three receptor forms becomes more evident in terms of the population of microstates ( Fig . 5d ) . The statistics of microstates shows that the receptor occupies different population of microstates depending on the type of ligand ( Fig . 5 ) . Among the entire microstates as summarized in Fig . 6 , ( i ) ≈ 80% of antagonist-bound form are populated in the 0000000000 or 0000000001 state . ( ii ) ≈ 20 . 2% of the switches in the agonist-bound form are in 1023th state ( 1111111111 ( 2 ) ) , and ≈ 31 . 28% are in 895th state ( 1101111111 ( 2 ) ) . ( iii ) Lastly , in the apo form of GPCR , on an average , 𝓢 1 , 𝓢 3 , 𝓢 10 are ON state , while 𝓢 2 , 𝓢 4 , 𝓢 6 , 𝓢 8 , 𝓢 9 are OFF state . Microstates that constitute the major population of the apo form are 1000000101 ( 2 ) ( 31 . 32% ) and 1001000101 ( 2 ) ( 10 . 56% ) . To quantify the statistical similarity explored by two different receptor states , say α and β ( α ≠ β ) in the 10-switch representation , we employ Hamming distance: d α β = ∑ i = 1 10 | ⟨ s α , i ⟩ - ⟨ s β , i ⟩ | ( 1 ) where ∣x∣ is the absolute value ( or modulus ) of x , and ⟨sξ , i⟩ is the average value of i-th switch ( i = 1 , 2… , 10 ) in the receptor form ξ , which is calculated in Fig . 5b . Since 0 ≤ ⟨sξ , i⟩ ≤ 1 , it is expected that 0 ≤ dαβ ≤ 10 . The more similar , the smaller the value of dαβ should be . We obtain dapo-ago = 4 . 74 , dapo-antago = 2 . 68 , dago-antago = 7 . 31 . Next , the “complexity” of each macrostate , defined with the ensemble of microstates {i = 1…Ns} where Ns = 210 , is quantified using the Shannon entropy: I = - ∑ i = 1 N s p i log 2 p i ( 2 ) where pi is the probability of occupying the i-th microstate as in Fig . 6 . When the receptor explores only a single state ( pk = 1 and pi ≠ k = 0 ) , the value of I should be I = Imin = 0; and if all the 210 states are uniformly explored ( pk = 2−10 for all k ) then I = Imax = 10 . Thus , the larger the value of I , the more diverse microstates are explored , which is also gleaned from Fig . 6 . We obtain Iantago ≈ 2 . 52 , Iapo ≈ 4 . 22 , Iago ≈ 3 . 70 , for antagonist-bound , apo , and agonist-bound state , resepctively; and hence the apo state explores the most diverse configurational space . Fig . 5e shows a schematic of relationship between the three receptor states combining the analyses using Eq . 1 and Eq . 2 . The apo form is more similar to the antagonist-bound form than to the agonist-bound form , which is consistent with the general notion that agonist contributes to the active GPCRs while both apo and antagonist contribute to the inactive state . Cross-correlations of the dynamics between binary switches . To identify the correlation between the ON/OFF dynamics of binary switches , we calculated their cross-correlation ( Cij ) by using the conformational ensemble from the simulations . C i j = ⟨ δ s i δ s j ⟩ ⟨ ( δ s i ) 2 ⟩ ⟨ ( δ s j ) 2 ⟩ ( 3 ) where δsi = si−⟨si⟩ is the variation of the switch value from its mean . Cij assesses the extent of coherence in the “change” in switch dynamics between the i-th and j-th switches . Marked differences of the correlation pattern are observed in the three distinct receptor states ( Fig . 7a ) : ( i ) The apo state ( the middle panel in Fig . 7 ) has only one positive correlation ( C 𝓢 4 𝓢 8 > 0 . 25 ) , and many other negative correlations ( C 𝓢 3 𝓢 4 , C 𝓢 3 𝓢 7 , C 𝓢 3 𝓢 8 , C𝓢1𝓢9<−0 . 25 ) ; ( ii ) By contrast , in the antagonist-bound state ( the left panel in Fig . 7 ) , a positive correlation ( > 0 . 25 ) is detected only between the 𝓢 3 and 𝓢 5 , and the negative correlations present in the apo state are suppressed; ( iii ) The agonist-bound state has a greater number of the positive correlations between the switches . The diagrams in Fig . 7b illustrate how the allosteric couplings are established among the switches , especially highlighting many positive couplings among switches in the agonist-bound state . Most notably , 𝓢 7 ( W246 ) , a central rotameric switch located at the deep bottom of ligand binding cleft , displays direct couplings with 6 other switches 𝓢 1 , 𝓢 2 , 𝓢 3 , 𝓢 5 , 𝓢 8 , and 𝓢 9 , and additional positive correlations are observed in C 𝓢 1 𝓢 2 , C 𝓢 2 𝓢 5 , C 𝓢 3 𝓢 8 , C 𝓢 5 𝓢 8 . This suggests that intramolecular signaling over the entire structure can be initiated by stimulating the 𝓢 7 . As can be confirmed from simulations , the indole 6-ring of W246 is within the range of hydrophobic interaction ( 4–5 Å ) [47] with the ethyl group of the agonist ( UK-432097 ) , while such direct interaction with W246 is missing in the antagonist ( ZM-241385 ) ( see Fig . 7c ) . Furthermore , there is a marked difference between the antagonist and agonist configurations in the orthosteric binding site; the antagonist is not stably poised as the agonist in the binding cleft ( Fig . 7c , middle panel and Fig . S7 ) . The dispersion of the center of position for each ligand is σantago = 2 . 204 Å and σago = 0 . 854 Å ( Fig . S7 ) . Although the residues at the binding site adopt diverse configurations , some key residues retain persistent interaction with the receptor ( Fig . S8a ) . Both the antagonist and agonist generally maintain the H-bondings with N253 and E169 residues ( Fig . S8b ) . Also , the adenine rings of the ligands interact with the phenyl ring of F168 via π-π interaction . Fig . S8c shows that the adenyl ring of the agonist interact directly with F168 while that of antagonist does not . Occasionally , the antagonist spins or flips at the binding site , indicating the lack of interaction with F168 residue . The polar residues located at the bottom of the binding pocket , i . e . , T88 , S277 , and H278 , only interact with the agonist . Especially , T88 , one of the hot spot residues in our earlier work [21] , maintains the stable H-bonding with the agonist . Along with stable configuration of the agonist ligand ( UK-432097 ) in the binding cleft ( See Fig . S7 ) , the widespread cross-correlation among the switch dynamics of 𝓢 1 , 𝓢 2 , 𝓢 3 , 𝓢 5 , 𝓢 7 , 𝓢 8 , and 𝓢 9 in the agonist form suggests that the interaction between agonist and W246 actuates allosteric signaling and contributes to a stable activation of the receptor function . Effect of inserting agonist to the apo state . Anticipating a detectable conformational change from inactive to active state , we generated 4 additional time traces by inserting agonist to the orthosteric binding site in the simulation trajectories of the apo state ( Fig . 8a , Fig . S9 . See Methods for the details of the simulation ) . The first two traces ( cases 1 & 2 ) were generated by inserting agonist at 125 ns and 150 ns when the ionic-lock was still intact ( s4 = 0 ) and were simulated for ≈ 750 ns . The second two traces ( cases 3 & 4 ) were generated at 595 ns and 625 ns when the ionic-lock was disrupted ( s4 = 1 ) , and were simulated for ≈ 250 ns . The consequences of the insertion of agonist , summarized in Fig . 8 , is still minor , which is evident when the average value of each switch ⟨si⟩ ( i = 1 , … , 10 ) is compared ( Fig . 5b middle panel versus Fig . 8b ) . This is not so surprising given that time scale of our simulation ( < 1 μsec ) is still too short to see a complete transition from inactive to active state in GPCRs , which is typically longer than milisecond time scale [48] . Although the overall trend of ⟨si⟩ looks similar , each trace from the insertion of agonist explores distinct microstate population ( Fig . 8c ) . Compared with the cases 1 & 2 , the 10 switch states were closer to those of agonist-state in the cases 3 & 4 when the agonist was inserted to the receptor with disrupted ionic-lock ( s4 = 1 ) ; in particular , the value of s4 for the cases 3 & 4 is greater and more variable ( Fig . 8a , b ) although the case 4 shows the stable rebinding of ionic-lock after 120 ns ( s4 = 1 → 0 , the rightmost panel in Fig . 8a ) . The complexity ( Eq . 2 ) of microstate ensemble explored in each simulation , are given in the table of Fig . 8d and the Hamming distances of the cases 1–4 from the three macrostates are calculated in Fig . 8e . The cases 3 and 4 are closer to the agoinist-bound state with dago-case3 = 4 . 11 and dago-case4 = 4 . 21 ( Fig . 8d , e ) than the cases 1 and 2 ( dago-case1 = 5 . 92 , dago-case1 = 6 . 60 ) . It is noteworthy that the binding of agonist to a receptor with stable ionic-lock ( cases 1 & 2 ) has driven the ensemble of microstates away from agonist-bound form and bring the ensemble close to the antagonist-bound state . Disruption of the ionic-lock in DRY motif is required for the activation of A2AAR .
Given that typical time scale associated with the transition from inactive to active GPCR conformers is ≳ O ( 1 ) msec [49] , it is practically impossible to capture the complete evidence of transition from inactive to active state using a single time trace lasting only 1 μsec [50] . Although the recent improvement in computational power and various computational strategies have ameliorated this time scale gap and have played significant roles in elucidating new facets of GPCR dynamics [31–33] , the gap of time scale between all-atom molecular dynamics simulation and experiments is still a serious problem in linking the computation of biomolecular dynamics with experimental observation [51] . Rather short in time scale ( ∼ 1 μsec ) , the trajectories from our simulations , which essentially probe the dynamics of receptor in terms of the noisy local variables , enable us to map the intramolecular signaling of three different ligand states . When it comes to the sampling efficiency of each macrostate , 1 μsec time scale is not too short to observe the individual dynamics of side chain flips that generally occur in picoseconds to nanoseconds timescale [52] , and we tried to decipher collective signals out of those noisy individual signals . The dynamic characteristics of the three macrostates are well discerned in consistent with the general notion of GPCRs , and our analysis based on the dynamics of 10 binary switches quantifies the differences among the thre statese . Our simulation results provide a comprehensive view of ligand-dependent conformational dynamics and show how the remote allosteric hotspots of A2AAR are coupled each other , which lend support on the pre-existing experimental data . For example , W246 has long been proposed to function as a central rotamer switch in the activation of GPCRs [5] . While the rotameric change of W246 was confirmed in spectroscopy [53] and some MD simulation studies [5 , 34] , X-ray crystal structures of A2AAR in both antagonist-bound and agonist-bound states show little difference of the position and dihedral angles of this residue . Our simulation captures the transition of the indole ring of W246 in the presence of agonist ( Supporting Movie M1 ) . In addition , our simulations clearly demonstrate characteristics of local dynamics of microswitches such as ligand-dependent state of ionic-lock , and the ligand-dependent variation of the rotameric angles in Y197 and Y288 [2 , 4 , 41 , 54] . Until recently , there are some other studies that have compared the conformational differences of A2AAR depending on the receptor states [34–36] . Pang et al . studied dynamic behaviors of A2AAR induced by the binding of two distinct antagonists [35] . Lee et al . simulated the ligand-dependent cholesterol interactions in A2AAR [36] . Exploring the distinct structural states that resemble the active and inactive states , Li et al . observed that the key structural elements change in a highly concerted fashion during the conformational transition [34] . Similar to our study , Li et al . highlighted the importance of the rotameric transition of W246 during the activation process . In contrast to these studies , which focused more on the ligand-receptor interactions , we tried to identify the communication among the microswitches . To best of our knowledge , our simulation is the first systematic study probing the dynamics of all microswtich residues of A2AAR , and made explicit that there are marked differences among the ligand-dependent conformational ensembles; and thus each ligand-dependent macrostate is made of mutually exclusive population of microstates , which supports the view that different type of ligands effectively remodel the protein energy landscapes [24 , 55] . In describing the conformational transition from inactive to active states in GPCRs , the recent studies by Yuan et al . [56] and Bhattacharya et al . [57] showed consistent and complementary results to our work , although the used methodologies were different . Yuan et al . discovered that a hydrophobic layer located inside of the helix structure forms a gate that opens to form a continuous water channel only upon the agonist binding [56] . The configuration of the nearby NPxxY motif affects this water channel , and the Gα peptide stabilizes the active state of Y7 . 53 which is one of the 10 switches in our study . Also , they calculated the average volume of the intracellular G-protein binding site and found that the volume is significantly increased upon the activation process , so that the intracellular site can accommodate G-protein binding . The volume change might well be a consequence of the swinging motion of TM5-ICL3-TM6 , and consistent with our result showing that the distance between the intracellular loop 2 ( ICL2 ) and ICL3 varies depending on the ligand binding states . Our study revealed the correlation between the ICL3 motion and the ionic-lock formation ( Fig . 2b ) . Bhattacharya et al . calculated the mutual information ( MI ) in internal coordinates from MD simulated trajectories , and they identified allosteric communication pipelines which are cenceptually similar to the long-range cross-correlation pathways discussed in our previous work [21] . They found that , in the inactive state , the allosteric pipelines mainly cross the TM6 helix , and as the state moves from intermediate to agonist-bound active state , these pipelines pass though the TM7 helix , suggesting that TM7 is increasingly correlated in the active state . As suggested in our study , diverse conformational space of GPCRs is dependent on the ligand binding states and is regulated by the allosteric pathways comprising of some hub residues . The works by these two groups and ours both surmise TM7 as the key helix in GPCR activation . While the most dynamic part is TM5-ICL3-TM6 region which accompanies the swinging motion , the hub residues , scattered throughout the helices , regulate the activation process in a concerted way . The antagonist-bound form explores the rotamer angle space entirely different from those in the agonist-bound and apo forms . The microstates visited by the apo form show an overlap with those by agonist-bound form although the extent of such overlap in the entire microstate population is vanishingly small and the duration of such overlap is only transient [58] . Occasional 10o-tilt of TM5-ICL3-TM6 relative to TM3 , shown in the simulation of the apo state ( Fig . 2b ) , may be related to the basal activities of the GPCRs . According to the pre-coupled theory , several class A GPCRs are suspected to be coupled to their corresponding G-proteins even in the inactive or basal states [59–62] . Inactive-state preassembly can facilitate the rapid and specific G protein activation[61] . Our simulation results suggest that the apo state of A2AAR can also form an inactive-state preassembly by visiting the microstates that overlap with those of the antagonist-bound state . We expect that a simulation of the apo form conducted in the presence of a G-protein will amplify this overlap and change the dynamics and correlation between the hotspots as well . Among several findings and predictions made in the present study , of particular note is the role of W246 ( 𝓢 7 ) in the GPCR activation . Although the importance of W246 residue were largely documented based on experimental or theoretical studies[5 , 34 , 53] , our cross-correlation analysis unequivocally indicates that W246 can work as a hub in the communications among the important microswitch residues ( Fig . 7 ) . Positioned deep inside the binding cleft , a signal of rotameric change of W246 , triggered by a direct hydrophobic interaction with an agonist , can be transmitted to the 6 other switches ( 𝓢 1 , 𝓢 2 , 𝓢 3 , 𝓢 5 , 𝓢 8 , 𝓢 9 ) that W246 is in direct correlation with . Our study confirms that W246 plays pivotal roles in GPCR activation as both an agonist sensor and actuator of allosteric signaling . The class A GPCRs are conventionally reported to function as monomers[63–66]; however , growing body of experimental evidence indicates that some GPCRs , including A2AAR , can form homodimers , heterodimers , or even higher level of oligomers[67–69] . In recent years , single-molecule imaging techniques revealed that GPCRs undergo dynamic equilibrium between monomers and dimers [70] , and studies of GPCR monomers and dimers are both meaningful and necessary . The dynamic features and correlations revealed in this study for monomer of A2AAR could change when the receptor forms dimers or higher level of oligomers . Thus , it would be interesting to investigate how dimerization of GPCR alters the correlations between hotspots . Faithful description of biomolecular dynamics , as a complex system , is a major challenge in both computational and experimental molecular biology . In this study , we try to simplify the description of each ligand-bound macrostate by selecting the 10 binary switches as the reaction coordinates . The conformational features of A2AAR captured in our 10 binary switch representation confirmed existing knowledge on the receptor and made specific predictions amenable to a further experimental study .
Preparation of the apo , antagonist- , and agonist-bound structures . The X-ray crystal structures of A2AAR bound with an agonist or antagonist were retrieved from the Protein Data Bank ( PDB ) . Since some loop regions are not resolved in these crystal structures , homology modeling was performed using the MODELLER program implemented in Discovery Studio v . 3 . 1 to prepare the full-length A2AAR models including all the loop regions . We used the mutation-free X-ray crystal structures of PDB IDs 3QAK [41] and 3EML [2] , which were available at the year 2011 we began this study , as the main templates for the agonist-bound and antagonist-bound models , respectively . To model the loop regions that were not determined in 3QAK and 3EML , the X-ray crystal structures with PDB IDs of 2YDV [4] and 3PWH [54] were used by retaining the conserved disulfide bridges connecting the loops , i . e . , C71-C159 , C74-C146 , C77-C166 , and C259-C262 . These models were optimized by a simulated annealing step and selected based on the Discrete Optimized Protein Energy ( DOPE ) score [71] . The final structures were energy-minimized with the Conjugate Gradient method and the Generalized Born with simple SWitching ( GBSW ) implicit solvent model [72] . We obtained the apo form by minimizing the receptor strucuture after removing the ligand from the antagonist-bound form . Simulations . By employing X-ray crystal structures of A2AAR and homology modeling to resolve the missing residues in the loop as described above , we performed MD simulations using NAMD package v . 2 . 8 with CHARMM22/CMAP force field . The topology and parameter files for the ligands were generated by SwissParam web server [73] . To construct the explicit membrane system , we first predicted the TM region of the A2AAR based on the Orientations of Proteins in Membranes ( OPM ) database , and subsequently surround the TM region with 173 ( apo ) , 182 ( antagonist-bound ) , and 177 ( agonist-bound ) 1-Palmitoyl-2-oleoylphosphatidylcholine ( POPC ) lipid molecules , which form a bilayer embedding the receptor ( length: 88 Å in x-axis , 91 Å in y-axis ) . The receptor in the membrane system was then solvated with 13 , 549 ( apo ) , 13 , 551 ( antagonist-bound ) , and 13 , 552 ( agonist-bound ) water molecules . We added 38 K+ and 49 Cl− ions to make ∼ 150 mM salt condition . As a result , the simulation system contains 68 , 799 ( apo ) , 70 , 052 ( antagonist-bound ) , and 69 , 449 ( agonist-bound ) atoms in 85 Å×88 Å×99 Å periodic box . Nonbonded interactions were smoothly switched off between 10 and 12 Å . To handle electrostatic interactions , the particle mesh Ewald algorithm was employed with a grid spacing smaller than 1 Å . The simulations were conducted via ( i ) energy-minimization of the initial system using the cojugate gradient method in the order of membrane , water molecules , and the entire molecules; ( ii ) gradual heating from 0 K to 300 K using a 0 . 01 K interval at each step; ( iii ) 50 nsec equilibration with NVT ensemble; and ( iv ) ∼ 1 μsec production run with NPT ensemble for each system with different ligands . No specific constraints , such as distance or angle , were applied during the simulations , except SETTLE algorithm for constraints in water molecules . We used the integration time step of 1 fsec , and for the analysis saved the simulated trajectories every 2 psec and sampled every 400 psec . To simulate the effect of inserting the agonist into the simulated apo-state , we first aligned the agonist ligand structure ( co-crystallized conformation in 3QAK . pdb ) to the apo-state structure , and then deleted the water molecules around 5 Å from the region where the ligand is to be inserted . After inserting the agonist , we performed the energy minimization and equilibrated the system for 20 nsec , so that water can solvate the ligand binding site as well as the agonist . The effect of agonist binding on the receptor structure was monitored subsequently . | As the key signal transmitters of a number of physiological processes , G-protein coupled receptors ( GPCRs ) are arguably one of the most important therapeutic targets . Orchestration of the intra-molecular signaling across transmembrane domain is key for the function of GPCRs . To investigate the microscopic underpinnings of intramolecular signaling that regulates the activation of GPCRs , we performed molecular dynamics simulations of the receptor in three distinct ligand-bound states using A2A adenosine receptor as a model system of GPCRs . Statistical analyses on the dynamics of and correlation among the 10 “binary switches” reveal that the three receptor states retain distinct dynamic properties . The antagonist- and agonist-bound forms of the receptors explore vastly different conformational space , and the apo form lies between them , yet located closer to the antagonist-bound form . In regard to the agonist-binding triggered activation mechanism , the correlation map among the 10 binary switches unequivocally shows that direct sensing of agonist ligand by the indole ring of W246 actuates the rest of intramolecular signaling . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Communication over the Network of Binary Switches Regulates the Activation of A2A Adenosine Receptor |
We used an individual-based molecular multisource approach to assess the epidemiological importance of Triatoma brasiliensis collected in distinct sites and ecotopes in Rio Grande do Norte State , Brazil . In the semi-arid zones of Brazil , this blood sucking bug is the most important vector of Trypanosoma cruzi—the parasite that causes Chagas disease . First , cytochrome b ( cytb ) and microsatellite markers were used for inferences on the genetic structure of five populations ( 108 bugs ) . Second , we determined the natural T . cruzi infection prevalence and parasite diversity in 126 bugs by amplifying a mini-exon gene from triatomine gut contents . Third , we identified the natural feeding sources of 60 T . brasiliensis by using the blood meal content via vertebrate cytb analysis . Demographic inferences based on cytb variation indicated expansion events in some sylvatic and domiciliary populations . Microsatellite results indicated gene flow between sylvatic and anthropic ( domiciliary and peridomiciliary ) populations , which threatens vector control efforts because sylvatic population are uncontrollable . A high natural T . cruzi infection prevalence ( 52–71% ) and two parasite lineages were found for the sylvatic foci , in which 68% of bugs had fed on Kerodon rupestris ( Rodentia: Caviidae ) , highlighting it as a potential reservoir . For peridomiciliary bugs , Galea spixii ( Rodentia: Caviidae ) was the main mammal feeding source , which may reinforce previous concerns about the potential of this animal to link the sylvatic and domiciliary T . cruzi cycles .
Blood-sucking bugs ( Triatominae , Reduviidae , Hemiptera ) are vectors of the parasite Trypanosoma cruzi ( Kinetoplastida , Trypanosomatidae ) , which causes Chagas disease . More than five million people are infected , and approximately 70 million live at risk [1] . An intensive and expensive Chagas Disease Control Program ( PCDCh ) —launched by the Ministry of Health of Brazil—was established in this country in 1975 directed primarily at Triatoma infestans—the vector responsible for most of cases of Chagas disease transmission in Brazil . As a result , in 2006 , T . cruzi transmission by T . infestans , had been officially interrupted [2] . Once it became clear that T . cruzi was no longer actively transmitted by T . infestans , investment in entomological surveillance and control programs decreased . As a consequence , concerns are now high regarding the potential of native triatomines , such as Triatoma brasiliensis , for re-emerging hyperendemic Chagas disease foci in Brazil [3] . Multidisciplinary studies have demonstrated that T . brasiliensis s . l . is a species complex that forms a monophyletic group [4–6] . It includes two subspecies ( T . b . brasiliensis; T . b . macromelasoma ) and three species ( T . juazeirensis , T . melanica , T . sherlocki ) . A taxonomic key has been published recently [7] . Members of the T . brasiliensis species complex are widespread in Brazil , occurring in 12 states , primarily within the Caatinga and Cerrado biomes [8–10] . Each presents distinct epidemiologic importance , morphological characteristics , natural history traits , ecological requirements , genetic characteristics and dispersal abilities [11–18] . Triatoma b . brasiliensis ( hereafter , T . brasiliensis ) is the best-studied member of the complex with respect to epidemiology and spatial distribution . It also has the highest prevalence of T . cruzi infection and a remarkable capacity to colonize homes . Therefore , T . brasiliensis is now the most important Chagas disease vector in the semi-arid zones of Brazil’s Caatinga ecoregion [17–18] . The primary difficulty with controlling T . brasiliensis is attributed to its capacity to occupy the domiciliary , peridomiciliary , and sylvatic environments [8] as it can readily re-infest treated buildings . Indeed , the actual foci for T . brasiliensis infestations are unknown: it was still unclear whether sylvatic foci are the major source of infestation or if bugs that invade domiciles are composed of domestic populations . To date , population genetic studies on T . brasiliensis have been based on a single mitochondrial gene , cytochrome b ( cytb ) [4 , 19] . But inferences using only cytb are limited due to its maternal inheritance and limited variation for small geographic scales . The properties and behaviors of individuals may determine their function within the systems they compose , and therefore we used an individual-based approach directed at individuals of T . brasiliensis collected in distinct sites and ecotopes . Using this approach , we first explored the genetic structure of T . brasiliensis by using both cytb and microsatellites markers [20] . We then estimated the natural prevalence of T . cruzi infection and the parasite diversity in bugs and identified natural feeding sources for T . brasiliensis . Finally , we were able to make inferences about the potential T . cruzi reservoirs . We suggested that Kerodon rupestris ( Rodentia: Caviidae ) may be a potential parasite reservoir , because most sylvatic bugs had fed on this rodent and were also infected with T . cruzi , Moreover , population genetics results indicated that vector control efforts in domiciles are threatened by gene flow from the perennial sylvatic foci with high T . cruzi prevalence .
Triatomines were collected in the municipality of Caicó in northeastern Brazil ( between 06°23’ to 06° 41’S and 36°58’ to 37°12'W; Table 1 ) , within the Caatinga ecoregion . The Caatinga is a mosaic of xerophytic , deciduous , semi-arid , thorny scrub . The study area chosen for insect captures was determined based on unpublished reports of public health institutions in charge of vector control , which indicated a combination of high pressure of infestation by T . brasiliensis after control activities with records of bugs infected by T . cruzi captured in domiciliary units . All investigated sites had not been sprayed in the three years prior the captures ( March ) . All captures were conducted during the rainy season . We sampled five sites for bugs: Caicó , São Fernando , São João do Sabugi , Santana and Solidade . At each site , we collected samples in the three ecotopes where this species is found: 1 ) domiciles ( Dom ) , or the indoor spaces of homes where triatomines are generally found in the crevices of mud walls , in furniture and under beds; 2 ) peridomiciliary areas ( Per ) , or the areas outside but within approximately 100 m of homes where domesticated animals sleep or are maintained; and 3 ) sylvatic areas ( Syl ) , or areas that are separate from peridomicilary areas and not occupied by humans . We maintain the traditional “sylvatic” terminology . However , we further divided sylvatic areas into “Syl-d”—degraded areas—where domesticated animals can be found ( sites F and S; see Table 1 ) , —and “Syl-c”—for conserved areas . These conserved areas belong to armed forces and are surrounded by huge halls . They are used only by soldiers during combat training . These areas are also protected from hunting . Domestic animals ( mainly cats ) are sometimes found in Syl-c , but they are removed in order to keep native fauna preserved . In domiciliary units , bug searches were conducted as routinely employed by local vector control-surveillance staff . We investigated all ecotopes usually colonized by T . brasiliensis in peridomiciles ( e . g . storerooms , henhouses , corrals , pigsties , and piles of tiles , bricks or stones ) and domiciles ( e . g . under stoves , beds , between spaces of roof , behind pictures fixed on the wall , in stored stuff , as food , home materials ) and manually captured with tweezers all bugs observed . In sylvatic environments , similar manual methods were employed; however , insects were searched with flashlights during the night in rocky outcrops . Domiciliary and peridomiciliary populations were defined according to the villages , in which bugs from three houses were sampled . Sylvatic populations were defined according to the rocky outcrop . In the study area , these formations were ~50 square meters . We obtained permission from house owners/residents to collect insects from all homes and properties . We identified adult vectors to species according to the taxonomic key [7 , 21] . Nymphs can only be identified molecularly because there are no morphological keys currently available . All the samples used in this study were also identified at the molecular level using cytochrome b marker ( see below ) . To identify whether insects were carrying T . cruzi-like parasites , we began with the traditional method used by local control-surveillance systems ( Rio Grande do Norte state's Health Department ) : one fecal drop from each bug was obtained by abdominal compression , then diluted in saline solution ( approximately 50 μl ) and examined fresh by microscopy at 220-400X . Because starved insects do not eliminate enough fecal drops , the success of this approach depends on the insects’ nutritional status and 90 insects could be examined . The insects used for molecular studies were placed in absolute ethanol . Only adults and fifth instar nymphs ( N5 ) were used for population genetic analyses to decrease sampling-strategic bias . Because older stages exhibit higher mobility , this strategy decreases the probability of sampling several relatives . Additionally , only populations with N ≥17 were used to ensure robust inferences . The resulting dataset for insect studies on population genetics was derived from 108 individuals distributed among five populations . Following previous study [22] , two sets of DNA extractions were performed for 126 insects with the DNeasy Tissue Kit ( Qiagen ) : the first came from legs ( DNA L ) and the second from the digestive tract ( DNA DT ) . By using DNA L , we amplified a portion of the insect cytb gene with the CYTB7432F/CYTBR primers [23 , 24] , ( Table 2 ) . DNA DT was used to identify both the feeding sources , by direct sequencing cytb with the vertebrate-specific primers L14841/ H15149 [25] and natural T . cruzi infection , by using a pool of three primers ( TCI/TCII/TC ) that amplify the non-transcribed intergenic region of T . cruzi mini-exon gene . TCI is specific to the T . cruzi I strain and TCII for T . cruzi II strain , generating a fragment of 300 and 350 bp , respectively [26] ( Table 2 ) . A DNA DT from Rhodnius prolixus experimentally infected with T . cruzi ( CL strain ) was used as a positive control . The negative control was purified water . Given the putative epidemiologic threat from domiciliary vectors , the T . cruzi natural identification and characterization was also conducted on small sample sizes ( N≤17 ) . For cytb amplification , PCR were performed with 25 cycles ( 94°C 60 s , Tm 60 s , 72°C 240 s; and for T . cruzi identification with 27 cycles ( 94°C 60 s , Tm 30 s , 72°C 30 s ) ( Table 2 ) . All amplifications were done on a thermal cycler Biometra T-1 Termoblock ( Germany ) . The amplified products were observed in 1% agarose gel , stained with ethidium bromide . PCR products were purified using a purification method with exonuclease I ( exoI ) and shrimp alkaline phosphatase ( sAP ) to degrade excess primers and nucleotides . PCR products were mixed with 1 . 5 μl of phosphatase buffer , 0 . 19 μl of exoI , 0 . 45 μl of UsAP ( USB , Cleveland , USA ) and incubated for 1 h at 37°C . The enzymes were inactivated for 15 min at 72°C . The DNA samples were quantified on Nanodrop 2000c Spectrophotometer ( Termo Fisher Scientific Inc . ) . All primers used and annealing temperatures for each PCR are given in Table 2 . Sequence reactions were performed using Bigdye Terminator v3 . 1 Matrix Standard Kit ( Applied Biosystem , UK ) and sequences were obtained using an automatic DNA sequencer ( ABI PRISM 3130 Genetic Analyzer sequencer , Applied Biosystem , UK ) . An initial alignment of each partial gene was conducted with the ClustalW2 algorithm [27] , and manual adjustments were made using SeaView v4 . 3 . 0 [28] . To identify the feeding sources and T . cruzi strains sequences were compared to those from the GenBank database using BLAST procedures . We only selected the best hit for each search using the cut off values of 95% query coverage and 10−100 E -value . The name of the putative host was retained at the specific level when the sequence identity was higher than 95% , but only at the genus level when it was in the range of 75–95% especially for the cold-blooded animals because sometimes they do not have an expressive representation in GenBank . For microsatellite genotyping , 108 adults and N5 were used , belonging to A ( N = 22 ) , B ( N = 17 ) , C ( N = 23 ) , D ( N = 27 ) and F ( N = 19 ) populations . The available primers for seven T . brasiliensis loci [20] were tested . We also revisited all the sequences with microsatellite inserts from our previous microsatellite-enriched genomic library and designed new primers for three loci ( Tb8102 , Tb 8150 , Tb2146 ) using OLIGO software ( version 4 . 0; National Biosciences Inc . ) . For each primer set , the optimal annealing temperature was determined using a gradient PCR ( MJ Research ) . In total , eight microsatellite loci were used in this study , because some previously designed [20] were monomorphic for our dataset ( Table 2 ) . PCR amplifications were performed using dye-labeled primer ( Applied Biosystem ) and for 40 cycles ( denaturation at 94°C for 30 s , annealing for 30 s , and extension at 72°C for 30 s; Table 2 ) . Microsatellite data were collected on an ABI Prism 3100 ( Applied Biosystem ) and alleles analyzed using GeneMapper ( Applied Biosystem ) , according to previous study [20] . We used two procedures to confirm the specific status of T . brasiliensis . First , we applied the Automatic Barcode Gap Discovery ( ABGD ) method [29] , using Kimura’ genetic distances , gap width X = 1 , and a set of prior minimum genetic distances ranging from 0 . 001 to 0 . 1 ( http://wwwabi . snv . jussieu . fr/public/abgd/abgdweb . html ) . Second , Bayesian inference ( BI ) was used to reconstruct phylogenetic relationships among unique haplotypes using MRBAYES v3 . 1 . 2 [30] , with the best-fit substitution model for the data under BI through the Akaike Information Criterion ( AIC ) determined with jModelTest 0 . 1 . 1 [31] . We ran 2 , 000 , 000 generations using the MCMC algorithm with a burn-in period of 500 , 000 using a GTR+I+Γ model . The dataset was composed of the sequences obtained in this study in addition to some downloaded from GenBank for members of T . brasiliensis species complex [4 , 5 , 20] . Triatoma sherlocki was used as outgroup . For cytochrome b analysis , the following population genetic summary statistics were calculated for the same set of data ( 108 N5 and adults ) : number of haplotypes ( Nh ) , haplotypic diversity ( Hd ) ( Nei , 1983 ) , nucleotide diversity ( Pi ) , mean number of pairwise cytb differences ( theta ) and average number of nucleotide differences ( k ) [32] . The evolutionary relationships among mitochondrial genotypes ( genealogy ) were evaluated using Network version 4 . 6 [33] . Tajima’s D [34] and Fu’s FS statistics [34–35] were performed to detect departure from a standard neutral model at equilibrium between mutation and genetic drift . For raw microsatellite data , possible genotyping errors due to stuttering , short allele dominance , and null alleles were tested using MICROCHECKER 2 . 2 . 3 [36] . Linkage disequilibrium was tested between all pairs of loci overall and in each population using FSTAT [37] . Deviations from Hardy Weinberg equilibrium were tested at each locus within each population using ARLEQUIN . 3 . 0 [38] , and the global test was run in GenePop [39] . Inbreeding coefficients ( FiIS ) were estimated using FSTAT [37] . We compared populations using FST pairwise estimates computed with ARLEQUIN [38] . Bonferroni correction was used throughout all levels to account for multiple testing [40] . Isolation by distance ( IBD ) was tested with a Mantel test for matrix correlation between pairwise genetics using distance ( FST/ ( 1-FST ) ) [41] and pairwise geographical distances with 10 , 000 randomizations and a reduced major axis regression ( RMA ) to calculate the slope in the program IBDWS version 3 . 23 [42] . The distributions of the expected heterozygosity for each locus and population under both the Infinite Allele Model ( IAM ) and the Stepwise Mutation Model ( SMM ) were calculated using BOTTLENECK [43] . The probability for the sign tests ( ps ) for heterozygosity excess and for pw Wilcoxon test ( two-tailed for H excess or deficiency ) were calculated . Finally , Geneclass2 was used to select or exclude populations as the origins of individuals [44] .
Fifteen sites ( Rio Grande do Norte , Brazil ) were sampled: four in sylvatic areas , four in peridomiciliary and seven in domiciliary environments ( Table 1 ) . A few T . pseudomaculata ( N = 14 ) and T . petrochii ( N = 3 ) adults were collected and excluded from further analysis , after identification by morphology . We identified 81 adult insects as T . brasiliensis by morphology , and the remaining 216 immature insects were identified as T . brasiliensis based on our molecular analysis ( see below ) . We used 108 insects that were composed of N5 and adults in this approach . Using the ABGD ( Automatic Barecode Gap Discovery ) species delimitation tool , the distribution of genetic distances display two modes separated by a “barcode gap” between 0 . 04 and 0 . 07 that leads to partitions with 5 groups for all taxa of the T . brasiliensis complex analyzed . This was true for all substitution models used ( p , KP or Tamura-Nei distances ) . All of our samples were clustered in the T . brasiliensis main group with 100% Bayesian posterior probability ( S1 File ) . Cytochrome b sequences showed 30 variable sites , including 13 singletons , resulting in 28 haplotypes for the 106 adults and N5 ( Table 3 ) . Overall , haplotype diversity was 0 . 89 , ranging from 0 . 77 ( peridomiciliary population D ) to 0 . 92 ( domiciliary population B ) . Domiciliary population B had the highest number of singletons ( N = 7 ) , while the sylvatic population F had the lowest ( N = 3 ) . The three sylvatic populations had similar haplotype diversities ( population A , 0 . 85; C , 0 . 87; F , 0 . 90 ) . Overall nucleotide diversity was 0 . 0054 , ranging from 0 . 0051 ( populations A , C and D ) to 0 . 0069 ( population B ) . To explore demographic population events , the Fu’s FS statistics were estimated ( Table 3 ) . Three populations ( B , C and F ) showed significant negative Fu’s FS values , especially the sylvatic population C ( P = 0 . 009 ) , suggesting expansion events . No significant departure from the standard neutral model was observed by the Tajima’s D values . Eleven haplotypes were shared by at least two populations; two haplotypes had the highest frequencies ( 18 . 87%; 25 . 47% for H_2 and H_5 , respectively ) and occurred in almost all sites . Four haplotypes were shared by the sylvatic and peridomiciliary populations , while the domiciliary population shared only two haplotypes with the sylvatic populations and three with the peridomiciliary populations . The sylvatic population C shared the most haplotypes with the others; namely , H_6 and H_5 with sylvatic populations A and F , respectively . The peridomiciliary population D and domiciliary population B showed the most unique haplotypes ( 4 and 6 , respectively ) . The haplotype network ( Fig 1 ) was relatively shallow , with a maximum of four mutational steps from central haplotypes ( H_2 and H_5 ) . Eight microsatellites loci were used to perform population genetic analyses ( Table 4 ) . The sylvatic population F and domiciliary population B had the lowest mean number of alleles per locus ( 4 . 33 and 5 . 0 , respectively ) and the peridomiciliary population D and sylvatic population A had the highest ( 6 . 29 and 5 . 86 , respectively ) . Populations A and C had the lowest observed heterozygosity ( 0 . 45 and 0 . 47 ) and populations D and B had the highest ( 0 . 51 and 0 . 52 ) . Except for population A , significant departures from Hardy-Weinberg equilibrium were found in three loci and linkage disequilibrium was detected for one ( populations A , C and D ) to three ( population B ) pair of loci . With our sampling strategy ( N5 and adults ) , no significant heterozygote deficiency ( FIS ) was observed , except for population A . Under the SMM , three populations ( A , C and D ) showed a significant heterozygote excess with the two tests used , but this was not the case using the IAM ( Table 4 ) . Gene flow among populations was estimated using FST pairwise comparisons . All pairwise comparisons except for three ( A-D , A-F , B-D ) were significantly different ( Table 5 ) . The FST values showed that the peridomiciliary population D is not differentiated from the domiciliary population B ( FST = 0 . 00852 , p<0 . 0001 ) and the sylvatic population A ( FST = 0 . 00635 , p<0 . 0001 ) , suggesting genetic flow . The lack of differentiation between populations A and F ( FST = -0 . 00078 , p<00001 ) likely indicates that sylvatic populations belong to the same panmictic unit whereas sylvatic populations A and C are geographically related but genetically differentiated ( FST = 0 . 01513; P< 0 . 05 ) . The Mantel test revealed no significant isolation by distance ( IBD; Z = 205 . 5983 , r = 0 . 2359 ) among populations ( See S3 File ) . Genetic assignment ( GeneClass2 ) was used to select or exclude populations as origins of individuals ( Table 6 ) . The results showed that high percentages of individuals were correctly assigned to their original population , especially for population D ( 93% ) with lower correct assignments for populations C ( 80% ) and F ( 79% ) . However , for sylvatic population A and domiciliary population B , 29% of the individuals were assigned to another population ( Table 6 ) . Using light microscopy , we were able to successfully analyze fecal drops from a total of 90 specimens from three populations ( B , C , and D ) . Forty-six percent of insects were infected with T . cruzi-like parasites . The infection prevalence was highest in the sylvatic environment ( 83% , population A; 95% , population C ) . We analyzed 126 insects using mini-exon amplification . We detected lower infection prevalence using molecular procedures than we did with direct analyses for sylvatic populations ( population A: 70 . 8% and population C: 52% ) . However , we were able to detect the presence of T . cruzi parasites in population D that we did not detect with the direct method . Consistent with direct fecal laboratory observations , none of the bugs from domiciliary population B was found to be infected using the molecular approach . However , one adult female insect from domiciliary population P was positive for T . cruzi , which was not included in the Table 7 due to the low number of insects captured for this population ( N = 3 ) ( see Table 1 ) . Each mammal feeding source ( except goats ) detected using cytb sequencing was also infected by T . cruzi , and so each should be considered potential a T . cruzi reservoir . Almost all T . cruzi were TCII; however , two samples from sylvatic population C were classified as mixed TCI and TCII lineages , exhibiting both the 300 and 350bp bands ( Table 7 ) . Notably , the positive bug from domiciliary population P was of the TCI lineage . All the TCII samples were sequenced ( N = 36 ) resulting in only one haplotype ( KT364456 ) . The strain Tu18 ( AY367125 . 1 ) showed the best hit matching ( 99% identity , 4e-100 ) . A high proportion ( 8/19 ) of the bugs that fed on K . rupestris was infected with TCII , while one was a mix of TCI/TCII .
All adult samples identified as T . brasiliensis using the taxonomic keys [7 , 21] were confirmed by ABGD and phylogenetic studies . Caicó is the type locality for T . brasiliensis [21] , where other vector species—such as T . pseudomaculata and T . petrochii—occur in sympatry . Triatoma pseudomaculata is commonly found concomitantly with T . brasiliensis in peridomicilary and domicilary areas [8] , whereas T . petrochii is sometimes found in rocky outcrops in the sylvatic environment [21] . In our study , we found that T . petrochii and T . brasiliensis co-occurred in some sylvatic environments and on the same rock outcrop spot . Even the adults of T . petrochii are morphologically similar to T . brasiliensis , as they were once in synonymy [21] . They are now recognized as full species and placed in the brasiliensis subcomplex of the infestans complex [45] . This co-occurring reinforces the importance of using ABGD approaches to address the taxonomic status of T . brasiliensis , especially for nymphs , for which there are no available taxonomic keys . Domiciliary population B had the highest haplotype diversity , revealing high genetic variability . We found that two central cytb haplotypes and other low-frequent haplotypes ( differing in one or two nucleotide substitutions ) formed a star-like haplotype network , which suggests population expansion . Even though D Tajima indicators were not significant , Fu’s FS showed that domiciliary population B and sylvatic populations C and F have more likely experienced a recent expansion event . Fu’s FS may be a more sensitive indicator for this event [34–35] than the Tajima D indicator , explaining this discrepancy . At most , four mutational steps from the central haplotype may be discerned in the network . Little structure using cytb haplotypes was observed , consistent with previous studies [19] for micro-scale geographic sampling . We used FST to infer the gene flow between populations . A recent study [46] has shown that for inferences based on FST by microsatellites in diploid organisms , results are robust even with minimal sample size ( N = 20 ) . The authors stated that improving the sampling design may show better results than genotyping more than 20 samples . With our sampling design , two populations did not meet the minimum sample size ( N = 17 and 19 individuals genotyped ) . Microsatellites applied to T . brasiliensis revealed low but significant genetic differentiation ( FST values ) between some neighboring populations , such as between sylvatic populations A and C , which are separated by less than 1 km . This differentiation may be associated with demographic events—low heterozygosity in population C might be related to bottlenecks . According to Luikart and Cornuet [47] , the SSM model is robust and suitable for microsatellites . It indicated possible bottleneck events for peridomiciliary population D and for sylvatic populations A and C . Similarly , the heterozygote deficit detected in sylvatic population A using FIS may be a result of the Walhund effect , reflecting the sampling of subpopulations with distinct allelic frequencies . It is worth noting that this sylvatic population A exhibited high indices of assignment outside of its own cluster , also suggesting that such subpopulations exist . The low variability in sylvatic population F , with both few alleles and low observed heterozygosity , may be a result of ecologic factors because it was collected in a degraded sylvatic area . In the Bolivian Andes , observations of the genetic differentiation between sylvatic and domiciliary populations of T . infestans were used to suggest that adjacent sylvatic foci are not the source for re-colonization , but rather , survivors of insecticide treated houses [48] . In Argentina , [49] a previous study showed that T . infestans populations from communities under sustained vector control were highly structured; however , populations from communities with sporadic vector control did not show such structure . Despite the great versatility in re-infestation profiles exhibited by T . dimidiata on the Yucatan Peninsula a variety of foci for re-infestations were found , which were depending on the site and the season [50] . Each of these examples were supported by the use of microsatellites . Our results support the idea that re-infestation foci for T . brasiliensis also come from distinct sources: peridomiciliary and sylvatic areas . This finding is worrisome from a vector control perspective because sylvatic populations represent perennial and uncontrollable foci for re-infestation . Additionally , domiciliary population B and peridomiciliary population D—which came from distinct villages—were genetically related , being probably in gene flow . The direction of re-colonization of homes after insecticide spraying is likely to be from peridomiciliary to domiciliary environments [51] . We reported 83–95% ( T . cruzi-like ) and 52–70% of T . cruzi prevalence via traditional and molecular methods for sylvatic populations of T . brasiliensis , respectively . The State Health Department in Rio Grande do Norte ( RN ) uses direct microscopic observation of fecal drops in Chagas disease control surveillance campaigns to identify T . cruzi . Using this method , we detected higher prevalence of T . cruzi infection in sylvatic populations than using molecular methods that might be a result of misidentifying T . cruzi-like parasites for T . cruzi during direct microscopic examination . Such misidentification might be also generating some proportion of false-positive results for local control-surveillance . There is also the possibility of false negative results using molecular analyses , due to PCR inhibition and/or annealing troubles . To prevent such results , a combination of other markers should be used . However , for domiciliary and peridomiciliary areas , infected bugs were only detected via molecular methods . False negatives acquired using traditional methods of direct fecal observations might be a result of low parasitic loads in insects . Overall infection prevalence for peridomiciliary and domiciliary environments in the entire Rio Grande do Norte state previously reported [8] was 4 . 3% . For both methods , our results contrast with some studies performed in Ceará State , where the T . cruzi infection prevalence in wild populations was 10 . 9% , with its highest peak in July [55] . On the other hand , a previous study [56] recorded higher rates in peridomiciliary ( 15 . 7% ) and domiciliary ( 10 . 7% ) areas than we identified in our study ( ≤ 3 . 6% ) . In this same area , previous studies [12] also noted 15 . 1% infection prevalence in Caico , without specifying the ecotope . The division of TCI and TCII proposed by authors [26] in 1996 was based on the two major phylogenetic lineages of T . cruzi . Therefore , despite being outdated and using lower resolution to define T . cruzi genotypes according the new consensus for T . cruzi nomenclature of Discrete Typing Units ( DTU ) [57] , our protocol exhibits an evolutionary meaning . All TCII we found based on our protocol [26] can be any of DTU II to VI , whereas TCI can be only DTU I , according to the new consensus for T , cruzi intraspecific nomenclature [57] . Indeed , by using this protocol [26] , we aimed mainly to confirm T . cruzi infection and to provide clues about the diversity in the area , according to these two major parasite lineages ( TCI and TCII ) . The TCII affiliation was confirmed by the similarity of the sequences with the Tu18 TCII strain . We found TCII prevalence ( 94% ) similar to that found previously [58] for this same municipality . Furthermore , their study [58] had the resolution to identify also the DTU III ( or TcIII ) , which was not targeted in our protocol . However , because TCII was prevalent in our study , further investigation is need to determine to which samples DTU corresponds , according to the new T . cruzi nomenclature . [57] . Our analysis found two ( 6% ) TCI isolates that correspond to DTU I [57] in the sylvatic area . These findings are consistent with previous studies [58–59] conducted in the same region and confirms that sylvatic T . brasiliensis population may harbor a mixture of the two major T . cruzi lineages . By detecting feeding sources and natural T . cruzi infection in the same T . brasiliensis individual , we can infer potential reservoirs . Except for the peridomiciliary population , T . brasiliensis fed mainly on mammals , which might be both the vector feeding sources and the T . cruzi reservoir . Differences in natural T . cruzi infection prevalence according to the ecotope can also be explained by the differences in feeding sources . Because birds are refractory to T . cruzi infection [21] , the natural infection was low in peridomiciliary areas where chickens were the dominant feeding source . In the sylvatic environment , no birds were identified as feeding sources . Instead , rodents were the major blood source—especially K . rupestris . This rodent probably plays an important role in the epidemiological T . cruzi cycle , as we found that it was the prevalent feeding source in the same T . brasiliensis individuals that were infected by T . cruzi . This suggestion should to be critically investigated , because the infected insects may have acquired T . cruzi via any other mammal before subjected to the test . Because cats and goats also served as feeding sources for insect vectors in sylvatic and domiciliary ecotopes , they may also link sylvatic and domiciliary T . cruzi cycles . Goats were identified recently as feeding sources for T . brasiliensis [60]; however , little is known of their role in Chagas disease epidemiology in northeastern Brazil . Here , we show a high frequency of feedings on cats in both sylvatic and domiciliary environments , including in an area with high natural T . cruzi prevalence . In Argentina [61] , cats are recognized as able to maintain T . cruzi populations . Indeed , cats are a common feeding source of triatomines almost everywhere [62] . Therefore , we recommend further study on the role of cats in T . cruzi transmission in northeastern Brazil . Using an individual-based multisource approach , we identified potential key-links for Chagas disease eco-epidemiology between the ecotopes examined . Microsatellites showed triatomine gene flow between ecotopic populations and highlighted varied sources of domiciliary and peridomiliciary infestations , which threaten vector control efforts . The co-occurrence of two T . cruzi strains in a sylvatic T . brasiliensis population is also consistent with a link between sylvatic and domiciliary cycles . As G . spixii circulates between distinct ecotopes , the finding of feedings on this mammal reinforces [60] the need to clarify the potential of this animal to connect T . cruzi from sylvatic to domiciliary environments . The distances between sylvatic and peridomestic spots where feedings were detected on G . spixii were less than 1 km . Therefore , the identification of this rodent as vector feeding source is even more relevant . Recently some authors [63]—testing a set of a priori hypotheses about the drivers of site-occupancy by T . brasiliensis—showed that T . brasiliensis was strongly associated with key hosts , including rodents in peridomiciliary environments . In these findings the ecotope structure had small to negligible effects . Our study did not characterize completely the dietary habits of T . brasiliensis because the rate of determination via direct sequencing was limited . In order to have a better representation of the host diversity , metagenomics or metabarcoding approaches using high throughput sequencing ( HTS ) technology is currently being developed . Additionally , to improve host assignation , the set of sequences deposited in GenBank , including for the fauna of neglected and remote sites , must likewise be improved . This limitation may have resulted in the low BLAST identity for some taxa not well studied . To exemplify this limitation , a sequence matched ( with 79% of identity ) an endemic lizard species ( Mabuya spinalis ) from Cape Verde Islands ( Africa ) [64] . This low identity may be a result of feedings on a related reptile species because there are records of occurrence of some other Mabuya sp . in northeastern Brazil [65] . Therefore , to improve host assignation , it will require the mobilization of taxonomists specialized on vertebrates and the constitution of a broader set of GenBank reference sequences . Because Chagas disease may be also transmitted through contact with hunted animals , as is suggested for the Amazon region in Brazil [66] , caution is necessary when handling these animals . Moreover , educational campaigns should address this issue . We recommend avoiding creating suitable habitats for G . spixx in peridomiciles [see 63] because it may connect sylvatic and domestic T . cruzi cycles . Chagas disease control is complex by nature due to an overlap of sylvatic , peridomestic and domestic cycles . However , the use of integrative approaches , such as population genetics to estimate gene flow between habitats combined with ecological data in a landscape genetics perspective [67–68] as well as an individual-based approach , linking vector population genetics to host and parasite identification , may help to better understand the eco-epidemiology of Chagas disease . | Blood-sucking bugs are vectors of the parasite Trypanosoma cruzi , which causes Chagas disease . Triatoma brasiliensis is the main Chagas disease vector in the Caatinga eco-region of northeastern Brazil . We showed that T . brasiliensis exhibits gene flow between sylvatic and anthropic ( domiciliary and peridomiciliary ) populations . This finding represents a challenge for vector control because sylvatic foci are not reached by control activities . Bugs from sylvatic foci harbor high T . cruzi prevalence ( 52–71% ) . The potentiality that the rodent Kerodon rupestris may act as a parasite reservoir was suggested because bugs infected by T . cruzi , were found in the same insects that fed on this animal . Peridomestic T . brasiliensis populations exhibited high prevalence for feeding on another sylvatic rodent , Galea spixii , which exhibit the potential to connect T . cruzi between ecotopes . The use of an individual-based approach directed to T . brasiliensis , linking vector population genetics to host and parasite identification , may help to better understand the eco-epidemiology of Chagas . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"population",
"genetics",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"animals",
"organisms",
"protozoans",
"neglected",
"tropical",
"diseases",
"population",
"biology",
"infectious",
... | 2016 | Molecular Individual-Based Approach on Triatoma brasiliensis: Inferences on Triatomine Foci, Trypanosoma cruzi Natural Infection Prevalence, Parasite Diversity and Feeding Sources |
The peritrophic matrix ( PM ) plays a key role in compartmentalization of the blood meal and as barrier to pathogens in many disease vectors . To establish an infection in sand flies , Leishmania must escape from the endoperitrophic space to prevent excretion with remnants of the blood meal digestion . In spite of the role played regarding Leishmania survival , little is known about sand fly PM molecular components and structural organization . We characterized three peritrophins ( PpPer1 , PpPer2 , and PpPer3 ) from Phlebotomus papatasi . PpPer1 and PpPer2 display , respectively , four and one chitin-binding domains ( CBDs ) . PpPer3 on the other hand has two CBDs , one mucin-like domain , and a putative domain with hallmarks of a CBD , but with changes in key amino acids . Temporal and spatial expression analyses show that PpPer1 is expressed specifically in the female midgut after blood feeding . PpPer2 and PpPer3 mRNAs were constitutively expressed in midgut and hindgut , with PpPer3 also being expressed in Malpighian tubules . PpPer2 was the only gene expressed in developmental stages . Interestingly , PpPer1 and PpPer3 expression are regulated by Le . major infection . Recombinant PpPer1 , PpPer2 and PpPer3 were obtained and shown to display similar biochemical profiles as the native; we also show that PpPer1 and PpPer2 are able to bind chitin . Knockdown of PpPer1 led to a 44% reduction in protein , which in spite of producing an effect on the percentage of infected sand flies , resulted in a 39% increase of parasite load at 48 h . Our data suggest that PpPer1 is a component for the P . papatasi PM and likely involved in the PM role as barrier against Le . major infection .
Leishmaniasis is a neglected vector-borne disease caused by several different species of Leishmania [1] , [2] . Current estimates of the distribution of leishmaniasis worldwide [3] indicate that endemic transmission occurs in 98 countries , with an approximate incidence of 500 , 000 new human cases are diagnosed annually , and 350 million people are at risk of becoming infected [4] . The DALY ( or disability adjusted life years ) burden for leishmaniasis is 2 million [4] . Leishmania are digenetic parasites , developing in a suitable mammalian host and within the sand fly vector [5] . To date , over 90 species of phlebotomine sand flies have been proven or incriminated as vectors of Leishmania [6] . In order to survive and successfully establish an infection in the sand fly , Leishmania must overcome many barriers ( reviewed by [2] ) . First , and following ingestion with the blood meal , transitional stage Leishmania amastigotes must survive a proteolytic attack by digestive enzymes [7]–[10] . Upon developing into the promastigote stage , parasites ( nectomonads ) escape from the endoperitrophic space after PM breakdown [7] , [11] and attach to the midgut epithelia [12] , [13] , in both cases to prevent excretion following the digestion of the blood meal . It has also been shown that an anterior plug prevents premature migration of nectomonads to anterior midgut [10] . As parasites develop into metacyclic promastigotes , they must detach from the midgut and migrate towards the foregut and the cardia ( or stomodeal valve area ) . At the cardia , it has been shown that Leishmania-secreted chitinase damages the stomodeal valve preventing its normal function , and forcing the sand fly to regurgitate the contents of the gut as it attempts to blood feed [14] , [15] . It is widely accepted that regurgitation carries Leishmania onto the skin of the vertebrate host , and this is the principal mechanism of parasite transmission . Regarding the sand fly PM , earlier findings suggested that it serves as a barrier against Leishmania development [16] , [17] . These results were further supported by feeding the chitinase inhibitor allosamidin to P . papatasi and showing that Leishmania major remained trapped inside a thicker PM [7] . These latter studies also revealed a dual role for the sand fly PM in protecting as well as serving as barrier to Leishmania . Altogether , these findings demonstrate that the PM is an important component of sand fly vector competence . Despite its importance , little is known about the molecular components of the sand fly PM [18] , [19] , or their roles during infection with Leishmania . Here , we characterized three peritrophins , PpPer1 , PpPer2 , and PpPer3 , previously identified in the midgut of P . papatasi [18] . PpPer1 and PpPer2 are likely involved in the formation of the PM scaffold , as suggested by their expression profiles and ability of the respective recombinant proteins to bind exogenous chitin . PpPer3 on the other hand may be involved in mechanisms related to protection of the epithelia , as this peritrophin displays a mucin domain and is expressed in both gut tissues and Malpighian tubules . We also investigated the role of the sand fly PM as a barrier for Leishmania development . Our results indicate that reduction of PpPer1 expression levels leads to an increase in Le . major load in P . papatasi . Altogether , our results suggest that PpPer1 plays a significant role in Le . major development within the vector midgut .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The use of animals in this study was reviewed and approved by the Committee on Institutional Animal Care and Use of the Kansas State University ( KSU-IACUC ) ( Permit Numbers 2747 , 2748 and 2749 ) . All sand fly feedings on animals and all bleeds were performed on animals under anesthesia , and all efforts were made to minimize suffering . The cDNA sequences of PpPer1 , PpPer2 , and PpPer3 were previously identified in [18] . Predicted isoelectric points and molecular weights of mature proteins were obtained using the Compute pI/Mw tool [20] . Putative secretory signal peptides were determined using SignalP 3 . 0 [21] . Prediction of O-linked glycosylated amino acids was carried out with NetOGlyc 3 . 1 [22] while N-linked glycosylation site prediction was performed using NetNGlyc 1 . 0 ( http://www . cbs . dtu . dk/services/NetNGlyc/ ) . Protein domains were identified by searching Prosite ( http://expasy . org/tools/scanprosite/ ) , Pfam ( http://pfam . sanger . ac . uk/search ) , and CDD ( http://www . ncbi . nlm . nih . gov/Structure/cdd/cdd . shtml ) domain databases . Chitin binding domain ( CBD ) classification in type-A ( CX13–20CX5–6CX9–19CX10–14CX4–14C ) , type-B ( CX12–13CX20–21CX10CX12CX2CX8CX7–12C ) , or type-C ( CX8–9CX17–21CX10–11CX12–13CX11C ) was performed visually , following the Consensus sequences described by Tellam [23] . The mucin-like domain amino acid composition was assessed using the GeneRunner software ( http://www . generunner . net/ ) . Predicted heme-regulatory motifs ( HRM ) were visually identified as cysteine-proline dipeptide [24] . Multiple sequence alignment of peritrophin CBDs was performed with the ClustalW tool in the BioEdit package [25] . Alignment was adjusted manually to remove some gaps . The CBDs of P . papatasi peritrophins were aligned to CBD sequences identified in peritrophins from Lutzomyia longipalpis [19] , [26] . Alignment was performed with each CBD sequence located between the first and sixth conserved cysteine residues . The L . longipalpis peritrophin cDNA sequence identified in whole body libraries ( NSFM-72d06 . q1k; [26] ) , referred to here as LlPer3 is an ortholog of the P . papatasi PpPer3 . A putative CBD was identified in the PpPer3 N-terminal sequence by visual inspection and named Pp3put . Ll3put was similarly identified within the L . longipalpis LlPer3 . Peritrophin sequences displaying similarities to Pp3put and Ll3put CBDs were retrieved from GenBank and aligned to the sand fly CBDs . Phylogenetic analysis was performed using the Maximum Likelihood method based on the Whelan and Goldman model [27] . The branch robustness was inferred by 500 bootstrap pseudo-replicates [28] . These analyses were carried out with the MEGA5 software [29] . All sand flies used were P . papatasi ( PPIS strain ) reared at Department of Entomology , Kansas State University , as previously described [30] . For the adult flies 3-to-5 day-old insects were used in the experiments described below . Blood feeding of sand flies was performed using two methods: 1 ) direct feeding on anesthetized ( 100 mg/kg of ketamine; and 3 mg/kg xylazine ) BALB/c mice; and 2 ) using glass feeders filled with heat-inactivated mouse blood or heat-inactivated mouse blood mixed with 5×106 Le . major amastigotes/ml [30] . Flies that were fed directly on the anesthetized mouse were used in the RT-PCR assays . The feeding of flies using the glass feeders was for flies used in real time PCR analyses , and for the flies injected with dsRNA . Only fully engorged sand flies were used . Phlebotomus papatasi dissections , RNA isolations , and cDNA syntheses were performed according to [30] . Total RNA obtained from various tissues from adult females were dissected and pooled as follows . For midguts , five tissues from sugar fed ( 0 h ) were dissected and combined . Likewise , five blood fed midguts also were dissected and combined at each of the following time points: 6 h , 12 h , 24 h , 36 h , 48 h , 72 h , 96 h , 120 h , and 144 h post blood meal ( PBM ) . Pools of adult carcasses , hindguts , heads plus salivary glands , ovaries , and Malpighian tubules were made from tissues obtained from single sand flies dissected at 0 h , 6 h , 12 h , 24 h , 36 h , 48 h , 72 h , 96 h , 120 h , and 144 h PBM . The pool of fat bodies was made from single flies dissected at 6 h , 12 h , 24 h , and 36 h PBM . RNAs from developmental stages were obtained from pools of 20 eggs , 10 L1 larvae , and five each for stages L2 , L3 , L4 , and five pupae . cDNAs obtained from RNA samples from pooled tissues were used in RT-PCR reactions described below . For the real-time PCR ( qRT-PCR ) assays , P . papatasi midguts were dissected from flies that fed either on blood or blood plus Le . major ( glass feeders ) at 24 h , 48 h , and 72 h PBM . cDNAs obtained from eight RNA samples representing individual midguts from each of the three time points were used for real time PCR reactions below . PpPer1 , PpPer2 , PpPer3 , and β-tubulin cDNAs were amplified using primer pairs described in Table 1 . The expression profiles of such genes were obtained after 23 amplification cycles for PpPer2 , 25 cycles for PpPer1 and PpPer3 , and 28 cycles for β-tubulin . Reactions were performed in 25 µl total volume , containing 12 . 5 µl GoTaq Master Mix ( Promega , Madison , WI ) , 1 µl cDNA , 0 . 5 pmoles each primer , and 10 . 5 µl molecular grade water . Amplification reactions were done as follows: 94°C for 3 minutes ( min ) ; 23–28 cycles of 94°C for 30 seconds ( sec ) , 57–58°C for 1 min , 72°C for 30 sec; and a final amplification step at 72°C for 10 min . Real time quantitative PCR reactions were performed with a MasterCycler Realplex4 Eppendorf Real-Time PCR ( Hamburg , Germany ) using BioRad SyBR green ( BioRad , Hercules , CA ) . Reactions were set up as described [30] . Amplification conditions and primer pairs used ( Table 1 ) were the same used in RT-PCR reactions , except that a total of 40 amplification steps were performed . As a housekeeping control , cDNA corresponding to the S3 protein of the 40S ribosomal subunit was amplified ( Table 1 ) . We assessed the expression of PpPer1 , PpPer2 and PpPer3 proteins present in midgut lysates . Pools containing five midguts dissected from non blood fed and from blood fed P . papatasi at 24 , 48 , 72 , 96 h PBM were homogenized in 50 µl PBS . Each midgut lysate was boiled in SDS lysis ( Invitrogen ) buffer for 5 min and one midgut equivalent from each midgut pool was separated under non reducing conditions on 4–12% Bis-Tris NuPAGE gels ( Invitrogen ) . Proteins were transferred to nitrocellulose and incubated for 16 h at 4°C with anti-PpPer1 and anti-PpPer3 antisera diluted 1∶100 in TBS-T . Blots were washed and incubated with anti-mouse conjugated to alkaline phosphatase ( 1∶10 , 000 ) for 1 h . Blots were developed using Western Blue ( Promega ) . Mature cDNAs sequences ( without the signal peptide ) corresponding to each of the three peritrophins PpPer1 , PpPer2 , and PpPer3 were PCR amplified from P . papatasi midgut using specific primer pair combinations ( PpPer1Mat_717F/PpPer1R-His , PpPer2Mat_219F/PpPer2R-His , and PpPer3Mat_867F/PpPer3R-His , Table 1 ) , with each of the reverse primers containing a 6×-His tag ( Table 1 ) . PCR amplifications were performed as follows: three cycles of 95°C for 3 min , 94°C for 1 min , and 68°C for 1 min; five cycles of 94°C for 1 min and 62°C for 1 min; and 25 cycles of 94°C for 1 min , 60°C for 1 min , and 72°C for 1 min . Each PCR product was subsequently cloned and purified as described [30]–[32] . Recombinant proteins were expressed using cells ( FreeStyle CHO-S ) and reagents obtained from Invitrogen , and according to the manufacturer protocols . Transfected cells were incubated at 37°C ( with 8% CO2 ) under gentle shaking ( 125 rpm ) . Culture supernatants were collected after 72 h and concentrated using Centricon filters ( Millipore ) at 3 kDa ( rPpPer2 ) or 10 kDa ( rPpPer1 and rPpPer3 ) cutoffs . Recombinant rPpPer1 and rPpPer2 also were produced using HEK293 cells and purified as described elsewhere [33] . The concentrated supernatants for rPpPer1 and rPpPer3 recombinant proteins were further purified by Ni-NTA HisTrap column ( GE Healthcare , Piscataway , NJ ) . Supernatants were filtered through a 0 . 22 µm Millex syringe filter ( Millipore ) , and dialyzed overnight in PBS at 4°C . After dialysis each supernatant was manually injected into a HisTrap column , previously equilibrated with binding buffer A ( 20 mM sodium phosphate buffer , 500 mM sodium chloride , 10 mM imidazole , pH 7 . 4 ) , and fitted to a HP1100 series HPLC system . Washes and elution of recombinant proteins were carried out according to [33] Different fractions corresponding to each wash interval with exception of the first 35 minutes were collected and analyzed by Comassie-stained SDS-PAGE and Western blot . For purification of rPpPer2 , a gravity flow column was used . The concentrated supernatant obtained from FreeStyle CHO-S cells was filtered and washed 5 times in PBS using a 3 kDa cutoff Centricon filter ( Millipore ) , and loaded onto 1 ml Ni-NTA column in a 5 ml syringe ( BD Biosciences ) . The column was washed with 15 ml 20 mM sodium phosphate buffer-300 mM sodium chloride-20 mM imidazole , and eluted with 5 ml 20 mM sodium phosphate buffer-300 mM sodium chloride-300 mM imidazole . The eluted rPpPer2 was concentrated ( 1 . 5 µg/µl ) and analyzed with Commassie-stained SDS-PAGE and Western blot . Native ( from P . papatasi midgut extract ) and recombinant proteins were subjected to HPLC gel filtration chromatography . The P . papatasi midgut extract was prepared from 10 midguts dissected 48 h PBM , pooled and homogenized in 50 µl of PBS pH 7 . 4 with 0 . 01% TritonX-100 followed by centrifugation at 14 , 000×g . The supernatant was collected and the pellet was extracted twice by sonication in 50 µl of PBS pH 6 . 8 , followed by centrifugation as above . The three resulting supernatants were combined . For the recombinant proteins , Ni-NTA purified rPpPer1 , rPpPer2 , and rPpPer3 were used . A 7 . 8×300 mm Bio-Sil SEC-250 column ( Biorad ) was fitted onto a HP Agilent 1100 series HPLC ( Hewlett Packard , Santa Clara , CA ) , and each protein was separately loaded onto the column . The HPLC was performed using PBS pH 6 . 8 at isocratic flow rate of 1 ml per minute . Absorbance of eluted was measured at 280 nm using the HP1100 series variable wavelength detector . One ml fractions of the HPLC elute ( both from midgut lysate or recombinant proteins ) were collected in 1 . 5 ml tubes and 10 µl of each fraction was blotted onto the nitrocellulose membrane ( dot blot ) . Dot blot filters were incubated to corresponding anti-sera and to anti-His antibodies as indicated below ( Western blot ) . Fractionation and dot blots were repeated twice for each recombinant protein and for the midgut lysate . Recombinants rPpPer1 and rPpPer2 obtained from HEK-293 cells also were fractionated using the HP1100 series HPLC , and collected fractions were applied to dot blots and incubated with anti-His or with specific antisera , as described above . N-linked deglycosylation was carried out with PNGase F ( New England Biolabs Inc . , Ipswich , MA ) . Briefly , 200 ng of purified rPpPer2 was incubated separately with 1000 , 2000 and 3000 units of PNGase F overnight at 37°C in 20 µl reaction . The reaction was stopped by addition of SDS gel loading buffer ( Invitrogen ) , the proteins were separated on 4–12% pre-cast Bis-Tris NuPAGE gel ( Invitrogen ) , transferred to nitrocellulose , and analyzed by western blot . We assessed the ability of the purified 6× His-tagged recombinant proteins to independently bind colloidal chitin according to [34] . Briefly , two micrograms of each recombinant protein was mixed with colloidal chitin suspended in 100 µl 10 mM sodium phosphate buffer , pH 8 . 0 , and incubated at room temperature for 1 h followed by centrifugation at 10 , 000×g . The supernatant was saved as unbound protein and the chitin pellet was washed with 50 µl of the same buffer and centrifuged as indicated above ( wash one or w1 ) . Additional washes were performed with 10 mM sodium phosphate buffer containing 1 M sodium chloride , pH 8 . 0 ( w2 ) and 0 . 1 M acetic acid ( w3 ) , respectively . The final pellet was boiled in 50 µl of SDS-PAGE sample buffer ( Invitrogen ) , centrifuged , and the supernatant collected . Unbound protein , washes , and SDS eluted fractions were separated in reducing condition with 2% betamercaptoethanol ( β-ME ) on pre-cast 4–12% Bis-Tris NuPAGE ( Invitrogen ) . Proteins were transferred to nitrocellulose and analyzed by Western blot . Antisera production was performed as described previously [31] , [32] , [35] . The mature sequences for PpPer1 , PpPer2 , and PpPer3 were amplified from P . papatasi midguts ( 12 h PBM ) , using the primer pairs PpPer1Mat_717F/PpPer1Mat_717R , PpPer2Mat_219F/PpPer2Mat_219R , and PpPer3Mat_867F/PpPer3Mat_867R ( Table 1 ) , respectively . Amplifications were as follows: 94°C for 3 min; 35 cycles of 94°C for 1 min , 56°C for 1 min , and 72°C for 1 min; and final extension at 72°C for 10 min . To determine the effects of dsRNA injection on the expression of PpPer1 in P . papatasi midguts , polyclonal anti-PpPer1 specific antisera ( 1∶50 dilution ) , as well as Western blot assays were performed according to [30] . Densitometry analysis was performed using the TotalLab TL100 software ( Nonlinear Dynamics , Durham , NC ) . Western blots also were performed for the analyses of the 6×His-tagged recombinant proteins rPpPer1 , rPpPer2 , and rPpPer3 . The concentration of each recombinant protein was determined using the BCA Protein Assay Kit ( Thermo Scientific , Rockford , IL ) or by measuring OD at 280 nm . Recombinant proteins ( purified or concentrated FreeStyle CHO-S supernatant ) were fractionated on 4–12% reducing Bis-Tris NuPAGE pre-cast gels ( Invitrogen ) . Proteins were transferred to a nitrocellulose filter ( Whatman , Dassel , Germany ) , incubated overnight at 4°C with anti-His antibody ( Santa Cruz , Santa Cruz , CA ) diluted 1∶2 , 000 or with different antisera , followed by three washes of 10 minutes each in TBS-T ( TBS buffer with 0 . 1% tween-20 ) . Each blot was incubated with anti-mouse antibody conjugated to alkaline phosphatase ( Promega ) diluted 1∶10 , 000 in TBS-T for 1 h at room temperature and washed in TBS-T as indicated above . The protein bands were visualized using the Western Blue ( Promega ) . Alternatively , Western blots were incubated with anti-mouse-HRP conjugated secondary antibody ( Promega ) diluted 1∶10 , 000 , and detected with SuperSignal West Pico Chemiluminescence Substrate ( Thermo Scientific ) in chemiluminescence assays . PpPer1 was selected for these studies in light of its mRNA expression profile ( midgut-specific and regulated by blood feeding ) . Double-strand RNAs were synthesized using the Megascript RNAi kit ( Ambion ) . Synthesis and purification of dsRNA as well as injections of sand flies were performed according to [30] . The dsRNA targeting PpPer1 ( dsPpPer1 ) was PCR amplified with primers PpPER1T7i_2 forward and reverse ( Table 1 ) . dsGFP was used as control , as described [36] . The effects of dsRNA induced knockdown were assessed by real time PCR analyses and Western blot . Next , we assessed the effects of PpPer1 knockdown on Le . major development within the P . papatasi midgut . In that case , 80 . 5 ng of dsRNA was injected intra-thoracically per sand fly [30] . After feeding on an infectious blood meal , midguts were individually dissected at 48 h and 96 h PBM and homogenized in 30 µl PBS ( pH 7 . 4 ) . Live parasites were counted using a hemocytometer . Unpaired t-test and Mann-Whitney U test were used to assess for statistically significant differences in expression profiles and parasite counts when data followed or not a normal distribution , respectively . Assessment of distribution pattern was carried out by D'Agostino & Pearson omnibus normality test . Differences were considered statistically significant at p<0 . 05 . All the statistical assessments were performed using GraphPad Prism v . 5 . 0 ( GraphPad Software , Inc ) . Sequence accession numbers: Sand flies . PpPer1 ( Gen Bank accession number: EU031912 ) . PpPer2 ( EU047543 ) . PpPer3 ( EU045354 ) . LuloPer1 ( EU124588 ) . LuloPer2 ( EU124602 ) . LuloPer3 ( EU124607 ) . LlPer3 ( AM093395 ) . C . felis . PL1 ( AAM21354 ) . C . quinquefasciatus . conserved hypothetical protein ( XP_001864216 ) . A . aegypti AaeL_AAEL012651 – Ae51put ( XP_001662775 ) , AaeL_AAEL012645 – Ae45put ( XP_001662776 ) , and AeputAaeL_AAEL012652 – Ae52put ( XP_001662772 ) .
The complete cDNA sequences of P . papatasi PpPer1 , PpPer2 , and PpPer3 were previously identified and published in [18] . PpPer1 open read frame ( ORF ) is 792 bp long , encoding a protein of 263 amino acids , with a predicted molecular weight of 28 kDa for the mature protein , and an acidic pI ( 4 . 84 ) . Putative N- and O-linked glycosylation at residues N29 and T211 , respectively , are expected to add to the molecular weight of the secreted protein . PpPer1 displays a predicted signal peptide ( amino acid residues 1–18 ) , suggesting the protein is secreted into the midgut lumen . Four type-A CBDs ( PpPer1 CBD consensus sequence: CX13–19CX5CX9–10CX12CX7C ) are also present in the mature protein . In addition , two putative HRM were identified at amino acid residues 182–183 in the third CBD ( Pp1CBD3 residues 44 and 45 in Figure 1 ) , and at residues 209–210 in the fourth CBD ( Pp1CBD4 residues 1 and 2 in Figure 1 ) . PpPer2 ORF is 270 bp coding for a predicted 7 . 8 kDa mature protein with a single type-A CBD ( consensus sequence: CX18CX5CX9CX12CX7C ) . Predicted N-glycosylation at amino acid residues N19 and N77 are also expected to increase the molecular weight of the protein . PpPer2 is an acidic protein ( pI 4 . 25 ) , with a single putative HRM at residues 22 and 23 of the predicted CBD ( corresponding to residues 1 and 2 of Pp2CBD1 in Figure 1 ) . The presence of a signal peptide with cleavage site between amino acids A17 and A18 suggests the protein is secreted into the midgut lumen . PpPer3 is 313 bp with two CBDs . Unlike PpPer1 and PpPer2 , PpPer3 has a mucin-like domain rich in serine ( 18 . 2% ) , threonine ( 36 . 4% ) , proline ( 12 . 1% ) , and glutamine ( 12 . 1% ) residues in addition to two type-A CBDs ( PpPer3 CBD consensus sequence: CX11CX5CX11–13CX12CX4–8C ) . The predicted molecular weight of the mature PpPer3 is 32 kDa . Moreover , this peritrophin has a neutral-to-basic pI ( 7 . 75 ) . As a large number of residues ( 281T , 285T , 286T , 292S , 293T , 294T , 295T , 296T , 300S , 301S , 302S , 303T , 304T , 305T , 306T , 307T , 309S , 310S ) within the mucin-like domain are predicted to be O-linked glycosylated , PpPer3 molecular mass is expected to be significantly greater . Two additional features of PpPer3 are the presence of a 57-residue long linker between the first ( Pp3CBD1 ) and second ( Pp3CBD2 ) CBDs , and an N-terminal sequence containing eight cysteine residues ( Pp3put , in Figure 1 ) . Although the Pp3put sequence displays a type-A CBD signature ( CX10CX5CX11CX14CX10C ) similar to other CBDs in P . papatasi peritrophins , it was not recognized as a bona fide CBD by standard bioinformatics' tools . Two predicted HRM were identified at residues 138–139 ( corresponding to residues 29 and 30 in Pp3CBD1 , Figure 1 ) and 150–151 ( residues 44 and 45 in Pp3CBD1 , Figure 1 ) in the first PpPer3 CBD while a single HRM was predicted in residues 262–263 ( residues 44 and 45 in Pp3CBD2 ) in the second PpPer3 CBD sequence ( Figure 1 ) . According to the multiple sequence alignment between P . papatasi and L . longipalpis peritrophin CBDs ( Figure 1 ) , the six conserved cysteine residues characteristic of type-A CBDs are present . Interestingly , the numbers of amino acid residues between the second and third cysteines , and between the fourth and fifth cysteines were the least variable . In addition , aromatic residues Y and F corresponding to positions 25 and 26 between the second and third cysteines , and position 48 between the fourth and fifth cysteines were detected . Regarding the HRM sites , most are co-localized with the first and fourth cysteine residues . Pp3put , a putative CBD domain present in the N-terminal portion of P . papatasi PpPer3 peritrophin , displays two extra cysteine residues at positions 18 and 53 , and two residue insertion ( PY ) between the fourth and fifth conserved cysteines ( Figure 1 ) . This putative CBD domain displays neither HRM motifs nor aromatic residues at positions 25 , 26 , and 48 . Interestingly , other insect peritrophins with features similar to Pp3put were also identified by searching the GenBank database against this putative CBD domain from P . papatasi . A phylogenetic analysis suggests a single clade for the Pp3put and Ll3put domains from sand flies ( P . papatasi and L . longipalpis ) , the Ae45put , Ae52put , and Ae51put from Ae . aegypti , the Cq16put from C . quinquefasciatus , and the CfPL1put from C . felis ( Figure 2 , blue box ) . The phylogenetic analysis also highlights the elevated conservancy that exists between the CBD domains in P . papatasi and L . longipalpis orthologous peritrophins , as reported previously [18] ) . The expression profiles of PpPer1 , PpPer2 , and PpPer3 were assessed by semi-quantitative RT-PCR ( Figure 3 ) . PpPer1 mRNA expression was adult midgut-specific and blood-induced ( Figure 3A and 3B ) ; transcripts were detected between 12 h and 72 h PBM , with the highest levels at 48 h PBM . PpPer2 transcripts were expressed in the midgut and in the hindgut ( Figure 3A and 3B ) . PpPer2 was constitutively expressed in sugar ( 0 h ) and blood fed guts ( Figure 3A ) . PpPer3 was expressed in the midgut ( Figure 3A ) and in the hindgut and Malpighian tubules ( Figure 3B ) . In spite of being expressed in sugar ( 0 h ) and blood fed midguts , PpPer3 mRNA expression was up-regulated between 12 h and 48 h PBM , somewhat similar to the PpPer1 expression profile . Among the three P . papatasi peritrophins , only PpPer2 was expressed in larval stages ( Figure 3C ) and , comparatively , also appeared to have the highest expression levels of the three peritrophins , according to Figure 3A . PpPer1 and PpPer3 are secreted in P . papatasi midgut following a blood meal , and are easily detected by Western blot at 48 h PBM ( Figure 4 ) . PpPer3 is also clearly present in midgut lysates dissected at 72 h and at 96 h PBM . Unfortunately , we were unable to resolve the smear visible in the lysates prepared from midguts dissected at 24 h PBM under non reducing conditions , and the specific antisera anti-PpPer1 and anti-PpPer3 did not bind to the native proteins under reducing conditions . The smear present in the 24 h samples is likely due to cross reaction of the antisera with mouse blood , and in our view irrelevant the findings . Nevertheless , at least for the native PpPer1 , its expression profile is in accordance to the mRNA expression profile observed in Figure 3A . We evaluated the effects of Le . major infection on expression of P . papatasi peritrophin mRNAs ( Figure 5 ) . PpPer1 expression displayed a statistically significant up-regulation ( 20% ) at 24 h post-infection ( Figure 5A ) . However , no statistical difference was observed for PpPer1 midgut expression at later time points ( 48 h and 72 h ) following Le . major infection . No difference was observed for PpPer2 for the three time points assessed ( Figure 5B ) . For PpPer3 , the mucin-like peritrophin , midgut mRNA levels were reduced by 25% at 24 h and 28% at 48 h after Le . major infection ( Figure 5C ) . No differential PpPer3 expression was observed in Le . major infected midguts at 72 h post-infection . Recombinant , 6×His-tagged , rPpPer1 , rPpPer2 , and rPpPer3 were successfully obtained using the FreeStyle CHO-S cells ( Figure 6A ) . Posttranslational modifications ( e . g . , glycosylation ) likely were responsible for the increased molecular weight detected for the recombinant proteins . Interestingly , mass spec analysis of rPpPer2 indicated a major peak at 9 . 2 kDa , with a minor peak at 18 . 4 kDa , suggestive of dimerization of this protein ( not shown ) . In contrast , the Western blot showed the presence of bands at approximately 16 kDa and 20-to-24 kDa in rPpPer2 ( Figure 6A ) . The higher molecular bands were no longer detected following digestion with PNGase F indicative of N-linked glycosylation ( Figure 6B ) . Differences between the predicted ( 7 . 8 kDa ) and the estimated sizes for rPpPer2 using mass spec and Western analyses might have been due to the presence of the 6×His tag interfering with gel migration , to non-predicted O-ring glycosylation ( s ) , or to incomplete digestion of N-linked residues . Similar SDS-PAGE migration discrepancies were observed for the peritrophin 15 of the screwworm fly Chrysomya bezziana [37] . The recombinant proteins were analyzed by gel filtration column . The retention time for rPpPer1 matched that detected for the native molecule present in the P . papatasi midgut lysate . For PpPer1 , elution from the midgut lysate occurred between fractions 6 and 7 ( dot blot 1 in Supporting Figure S1A ) , while the rPpPer1 was eluted with fractions 6–8 ( Supp . Figure S1B , dot blot ) . For rPpPer2 , the protein was eluted mainly in fractions 11–13 ( Supp . Figure S1C , dot blot ) . We were unable to generate polyclonal anti-sera to PpPer2 to efficiently detect fraction from the midgut lysate . Nevertheless , our data clearly demonstrates that the rPpPer2 was produced , as determined by hybridization with the anti-His antibody ( Supp . Figure S1C ) . For PpPer3 , while the recombinant protein had a wide trailing , being eluted with fractions 6–13 ( Supp . Figure S1D ) the native protein was eluted with fractions 6 and 7 ( dot blot 2 in Supp . Figure S1A ) . Figure 7 shows the results of the binding of the recombinant proteins to colloidal chitin . No recombinant protein in detected in the wash fractions obtained from the colloidal chitin binding assays ( washes 1–3 ) . Accordingly , rPpPer1 ( Figure 7A ) and rPpPer2 ( Figure 7B ) were only detected after boiling in the presence of SDS , thus demonstrating the ability of both rPpPer1 and rPpPer2 to bind chitin . In contrast , we were unable to demonstrate binding of rPpPer3 to chitin . PpPer1 was selected for the knockdown experiments following our assessments of its expression profile ( Figure 3 ) and according to the data from chitin binding assays . As PpPer1 is expressed exclusively in the midgut after blood feeding and binds chitin , we reasoned it was involved in PM formation . Intra thoracic injections of P . papatasi females with 80 . 5 ng of double-strand RNA specific for PpPer1 ( dsPpPer1 ) were performed to assess the role , if any , of PpPer1 protein on Le . major development . First we determined whether injection of dsPer1 was able to reduce mRNA and protein levels . As shown in Figure 8 , injection of the dsPpPer1 led to 45% and 30% reduction in mRNA expression levels at 24 h and 48 h PBM , respectively ( Figure 8A ) , and to a corresponding reduction of 44% in protein levels at 24 h PBM ( Figure 8B and C ) . In spite of the lack of noticeable differences in PM structure between dsPpPer1 RNA-injected versus control-injected sand flies following midgut dissection ( not shown ) , knockdown of PpPer1 led to an increase in Le . major load within P . papatasi midguts of 39% at 48 h and 22% at 96 h post-infection ( as shown in Figures 9A and 9B ) .
Here , we characterized three peritrophins of the sand fly P . papatasi thought to be involved in the formation of the PM in adult females . In addition , as the PM is an important component of vector competence in sand flies [7] , [16] , [17] , [30] , we assessed the role of PpPer1 as a molecular barrier against Le . major development . Different than PpPer1 and PpPer2 , PpPer3 is a mucin-like peritrophin with two CBDs and a mucin-like domain rich in serine , threonine , glutamine , and proline residues . Mucin-like domains are predicted to be heavily O-linked glycosylated that contributes to a gel-like consistency for the PM , critical to their role in protecting the midgut epithelia from abrasion , hydrolytic enzymes , heavy metals , and pathogens [38] , [39] . In addition to their role in PM formation , peritrophins are also known to participate in detoxification [40] . In the mosquito A . aegypti , the mucin-like peritrophin AeIMUC was shown to bind heme in vitro via heme-regulatory motifs ( HRM ) , while it also bound chitin [40] , [41] . HRMs are predicted for all three P . papatasi as well as the L . longipalpis peritrophins , suggesting a role for these proteins in heme binding and detoxification in sand flies . We identified the CBDs present in the three peritrophins from P . papatasi and those identified in L . longipalpis sequence databases as type-A CBDs displaying the molecular hallmarks required for chitin binding [23] , [42] , [43] . Chitin binding hallmarks include six conserved cysteine residues ( with a conserved number of residues between each conserved cysteines matching the consensus sequence for type-A CBDs ) [23] , and conserved aromatic amino acids residues predicted to interact with chitin fibrils [23] , [43] . Although the two putative CBDs Pp3put and Ll3put display the six conserved cysteines interspaced by the characteristic length expected for type-A CBDs [23] , amino acids other than aromatics are found in these putative domains . In addition , the Pp3put and Ll3put CBDs display two extra cysteines at residues 20 and 54 and have an unusual two-peptide insertion between the fourth and fifth conserved cysteines . Such features are in contrast with all the other bona fide sand fly type-A CBDs that display a conserved number of amino acids between the fourth and fifth cysteines ( 12 residues ) , and lack the extra cysteines present in the putative domains . Thus , it is tempting to speculate that Pp3put and Ll3put CBD domains underwent some type of neo-functionalization [44] , and that such change might also have occurred in peritrophins of other insects , such as the cat flea and mosquitoes . This hypothesis is to some extent supported by our findings that PpPer3 ( which has the Pp3put domain ) also is expressed in the Malpighian tubules , and by the lack of binding of rPpPer3 to chitin ( see below ) . Although the latter may be due to other factors , the expression in tissues other than the sand fly gut is suggestive of an additional function for this protein . Future functional characterization of these putative domains , both in P . papatasi and in L . longipalpis , will shed light on their functions . We also obtained recombinant rPpPer1 , rPpPer2 , and rPpPer3 and confirmed the ability of rPpPer1 and rPpPer2 to bind colloidal chitin and thus likely be involved in the formation of PM1 in P . papatasi . Our assessment of temporal and spatial expression of the P . papatasi peritrophins demonstrated that PpPer1 expression is midgut-specific and blood-induced , resembling the transcriptional profile of PpChit1 , a midgut specific P . papatasi chitinase [32] . The PpPer1 protein appears to be secreted in the midgut at an earlier time point ( 24 h PBM ) following a blood meal than PpChit1 , whose activity peaks between 48 h and 72 h PBM [30] . Nonetheless , these patterns of protein expression are consistent with the functional roles of PpPer1 and PpChit1 in PM formation and degradation , respectively . In contrast to the results observed for PpPer1 , PpPer2 and PpPer3 mRNAs are expressed prior to blood feeding ( constitutively ) , and their expression is not limited to the midgut: PpPer2 is expressed in the midgut and hindgut; while PpPer3 is expressed in the midgut , hindgut , and Malpighian tubules . Although peritrophin expression in hindguts and/or Malpighian tubules have also been detected in the cat flea Ctenocephalides felis [45] and the fruit fly Drosophila melanogaster [46] , the physiological roles of peritrophins in these tissues have not been determined . Regarding the constitutive expression patterns of PpPer2 and PpPer3 , the corresponding proteins might not be translated in the same fashion . In Aedes aegypti , the peritrophin AeIMUCI is constitutively expressed , and yet the protein is only detected in blood fed midguts and up to 24 h PBM [40] . PpPer2 also is expressed in larval stages , similar to AeIMUCI [41] . However , following our assessment of midgut lysates for native peritrophin expression , it became evident that at least for PpPer1 , protein expression in the midgut correlates with mRNA expression . Native PpPer1 is readily detected at 48 h PBM , matching its peak of mRNA expression . For native PpPer3 , we observed significant levels of protein at 48 h , and lower amounts at 72 h and 96 h PBM . This profile correlates at least partially with the mRNA expression profiles observed . To assess whether or not Le . major infection is capable of modulating P . papatasi peritrophin gene expression , we compared PpPer1 , PpPer2 , and PpPer3 mRNA levels between Le . major infected and blood fed midguts . Although Le . major infection was not able to modulate PpPer2 expression profile , PpPer1 and PpPer3 expression levels changed significantly upon infection . Regulation of peritrophins was suggested by previous transcriptome analyses studies of both P . papatasi and L . longipalpis [18] , [19] , [47] . The expression of PpPer1 was up-regulated at 24 h post-infection whereas PpPer3 mRNA levels were reduced at 24 h and 48 h post-infection . Up-regulation of PpPer1 by Le . major may assist in protecting the parasite against proteolytic enzymes ( parasite advantage ) , or may be a response by the sand fly in order to possibly reduce permeability of the PM ( disadvantageous to the parasite ) . Whether one or multiple signals secreted by the parasite or present in the infected blood are involved in this regulation still needs to be determined . Similarly , regarding PpPer3 , whether differential gene expression in infected midguts was parasite-mediated , or a vector-induced defensive response against infection , needs to be further investigated . The role of the P . papatasi peritrophin PpPer1 in Le . major development in P . papatasi midgut was assessed via RNA interference ( RNAi ) . PpPer1 was chosen for RNAi experiments because it was shown to be expressed exclusively in the midgut , and only after blood feeding . The sand fly PM is thought to fulfill two apparently opposing roles ( protection and barrier ) when it comes to Leishmania infection [2] . That sand fly PM protects Le . major against digestive enzymes early in infection [7] initially suggested to us that a potential alteration of the PM scaffold , increasing its permeability by the knockdown or removal of one or more peritrophins , would lead to killing of parasites . The injection of dsPpPer1 into P . papatasi thorax reduced PpPer1 mRNA and PpPer1 protein levels by 45% . Although no difference in the prevalence of infected sand flies was observed at 48 h post-infection ( not shown ) , PpPer1 knockdown led to a significant increase ( 39% ) in Le . major load in P . papatasi midguts . One possibility to explain these results is that a PM with increased permeability may allow a greater influx of digestive enzymes to the endoperitrophic space , turning blood meal digestion faster and making nutrients more readily available for Le . major multiplication . Although increased Le . major load in PpPer1 knockdown also was noted at 96 h post-infection ( 22% ) , this increase was not statistically significant . Our results showing that knockdown of PpPer1 with concomitant increase in parasite load at 48 h post infection do not exactly contradict those of Pimenta et al . [7] . In that case , killing of Leishmania was induced by the complete lack of the PM by addition of exogenous chitinases to the blood meal which might also have affected the parasite's own chitinase . In contrast , here , we specifically knocked down a single peritrophin but , overall , the PM was still present likely leading to the distinct outcome between the two studies: killing of Leishmania versus increased load . Future studies to assess the changes in PM porosity caused by the knockdown of PpPer1 and how such changes may affect the flow of digestive proteases and nutrients in and out of the endoperitrophic space will help clarify these issues . Recently , Araujo et al [48] showed that adding chitinase to the blood meal led to accelerated egg-laying by sand flies and a reduction of the total eggs laid per females , and it was likely associated with a speedier acquisition of nutrients due to the lack of the PM . Conversely , current data from our laboratory ( not shown ) indicate that feeding flies with antisera targeting the sand fly midgut chitinase PpChit1 [32] leads to a delay in egg laying after blood feeding and an increase the number of eggs laid per female . Taken together these results suggest that the integrity and the permeability of the PM interfere with the development of Leishmania as well as with fitness in sand flies . | For a successful development within the midgut of the sand fly vector , Leishmania must overcome several barriers imposed by the vector that include the digestive proteases secreted within the midgut following a blood meal by the insect , the need to escape from the endoperitrophic space , and attachment to the midgut epithelia to prevent excretion with the remnants of the blood meal . The sand fly peritrophic matrix ( PM ) constitutes an important barrier against the establishment of Leishmania within the sand fly and if trapped within the PM these parasites will be passed along with the remnants of the blood meal . Despite the role of sand fly PM on Leishmania development , characterization of its molecular components and assessment of their roles against Leishmania are lacking . Thereby , we performed the molecular characterization of three P . papatasi peritrophins named PpPer1 , PpPer2 , and PpPer3 . Overall , we demonstrated that: ( 1 ) PpPer3 displays a putative CBD domain that might have undertaken neo-functionalization , ( 2 ) PpPer1 and PpPer3 genes display differential gene expression upon Le . major infection; and ( 3 ) PpPer1 seems to be an important component for the function of P . papatasi PM as a barrier against Le . major infection . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"functional",
"genomics",
"molecular",
"cell",
"biology",
"biology",
"genomics",
"microbiology",
"genetics",
"and",
"genomics"
] | 2013 | Characterization of Phlebotomus papatasi Peritrophins, and the Role of PpPer1 in Leishmania major Survival in its Natural Vector |
Essential genes code for fundamental cellular functions required for the viability of an organism . For this reason , essential genes are often highly conserved across organisms . However , this is not always the case: orthologues of genes that are essential in one organism are sometimes not essential in other organisms or are absent from their genomes . This suggests that , in the course of evolution , essential genes can be rendered nonessential . How can a gene become non-essential ? Here we used genetic manipulation to deplete the products of 26 different essential genes in Escherichia coli . This depletion results in a lethal phenotype , which could often be rescued by the overexpression of a non-homologous , non-essential gene , most likely through replacement of the essential function . We also show that , in a smaller number of cases , the essential genes can be fully deleted from the genome , suggesting that complete functional replacement is possible . Finally , we show that essential genes whose function can be replaced in the laboratory are more likely to be non-essential or not present in other taxa . These results are consistent with the notion that patterns of evolutionary conservation of essential genes are influenced by their compensability—that is , by how easily they can be functionally replaced , for example through increased expression of other genes .
Essential genes code for central cellular processes required for the viability of an organism . Many recent studies have used experimental data to determine gene essentiality in a large number of bacteria [1]–[11] . The central role that essential genes play suggests they should be highly conserved during evolution , and several comparative genomic analyses have confirmed this hypothesis [12]–[15] . A second implication of this pattern of conservation is that essential genes tend to remain essential during evolution: if a gene is essential in one organism , then orthologues of that gene are usually essential in other organisms ( Figure 1 ) . However , there are many genes that do not follow these patterns: some genes that are essential in one organism are non-essential in other organisms; in other cases , genes that are essential in one organism are absent or have been lost from the genomes of other organisms [13] , [15] , [16] . Instances in which essential genes have become non-essential , or have been lost completely from genomes , suggest that either changes in physiological or environmental conditions have altered the essentiality of a gene , or that the genetic context has changed in a way to allow loss of a previously essential function . In this case , a second gene ( either paralogous or unrelated to the original essential gene ) may now perform the essential function . This raises the question of whether there is a connection between compensability and conservation level of essential genes . The possibility of a connection between compensability and gene conservation has been raised on at least one occasion previously . Geissler et al . [17] observed that the Escherichia coli cell division protein ZipA is poorly conserved in other taxa . Under the assumption that other proteins must fulfill this role in these other taxa , they looked for suppressor mutations that would obviate the requirement for ZipA , and found that a single mutation in FtsA suppressed the lethal zipA phenotype . Here we used a systematic approach to investigate how frequently the functions of essential genes of Escherichia coli can be replaced under laboratory conditions , and whether the frequency of this process correlates with patterns of evolutionary conservation . To gain insight into this question , we used the following methodology . First , we compromised the function of an essential gene in Escherichia coli by decreasing its expression with a tightly regulated promoter . We then increased the expression level of a large number of other genes to identify genes that are capable of compensating for the function of the impaired essential gene . Repeating this process for a large number of essential genes , we isolated a set of genes that can functionally replace the essential genes when overexpressed . We find that although the majority of these compensating genes are not homologous to the impaired essential genes , they exhibit similar functions . In a few cases , the compensating genes are capable of fully replacing the functions of the essential gene , allowing the deletion of the essential gene from the genome . Finally , we show that those essential genes whose function can be compensated for in the laboratory are more likely to be non-essential or not present in other bacterial genomes , raising the possibility that similar compensatory mechanisms may allow essential gene loss to occur in natural populations . Many previous studies have shown that gene essentiality is a mutable characteristic and is dependent on both the genetic background of the organism or the environmental conditions [18]–[21] ( such genes are termed conditionally essential genes ) . The results we present here imply that in some cases it may be possible to predict which essential genes are more likely to be conditionally essential .
We constructed a collection of Escherichia coli strains in which essential genes were placed under the control of a conditionally expressed promoter . The essential genes were selected from a variety of functional classes [22] , and exhibit a wide range of conservation levels [15] . In addition , some of these genes have been consistently found to be essential across all bacteria that have been examined empirically , while others are essential in only a few ( Figure 1 ) . A total of 26 genes were chosen ( approximately 10% of the essential gene content of E . coli ) ; six of these genes are in essential tandem operons ( nrdAB , dnaTC and glmUS ) . To control the expression of the essential genes , we replaced their native promoters with the arabinose-inducible araBAD promoter ( Para; see Methods ) . By shifting these mutant strains from medium with L-arabinose to medium without L-arabinose and supplemented with D-glucose , expression of the essential gene was repressed , and in all cases this resulted in growth inhibition or severe growth defects ( Figure 2 ) . We also tested if a plasmid encoding the corresponding essential gene rescued the lethal growth phenotype , and this was the case , except for the operons dnaTC and nrdAB ( Table S1 ) . This is most likely due to the fact that these are tandem operons , and transformation and maintenance of two separate plasmids that complement the function of each gene is unlikely , as the plasmids share the same replication origins and resistance markers . Next , we assembled a library of overexpression plasmids using the ASKA ( - ) strain collection [23] . The ASKA ( - ) library consists of 4123 clones that each contain a plasmid with one E . coli open reading frame cloned behind an IPTG-inducible Plac promoter and a N-terminal ( His ) 6-tag . Addition of IPTG induces strong expression of the downstream open reading frame . We removed the 26 clones containing plasmids with the essential genes mentioned above from the ASKA library . We then pooled the remaining 4097 clones and extracted plasmid DNA from this mixed pool . Each conditional lethal mutant was transformed with an aliquot of the purified plasmid pool and plated on restrictive agar plates ( where the essential genes were not expressed ) with IPTG to induce expression of plasmid-encoded genes . We induced expression of the plasmid encoded genes with 50 µM IPTG , because higher induction levels are known to be deleterious for growth: 51% of E . coli proteins expressed under control of Plac at 1 mM IPTG cause lethality [23] . We also measured transformation efficiency: when transformed and plated under permissive conditions with selection for the plasmid-encoded chloramphenicol resistance , all strains except murA and fldA gave rise to at least 1 . 2×104 colonies , which is the minimal number of clones required for coverage of 95% of all variants transformed [24] ( see Table S1 ) . We recovered up to 10 transformants from each plate , and restreaked them onto restrictive agar plates with IPTG to confirm growth . Upon successful growth of these clones , plasmids were extracted . To discriminate between possible chromosomal suppressor mutations and high copy suppressors ( HCS ) encoded on plasmids , each plasmid was retransformed into the ancestral conditional lethal mutant under permissive conditions , and colonies were tested for growth under restrictive conditions . In case of successful growth , the plasmid was sequenced . To control for suppression by multiple plasmids , we purified plasmids from the ASKA library and repeated the retransformation test , yielding the same results as before . For cases in which high copy suppression by the purified plasmid could not be confirmed , a second round of transformation and selection as described above was carried out . For six strains ( Para-aspS , plsC , plsB , ffh , glmUS , and gltX ) , although colonies were recovered , no HCS plasmids were isolated after two rounds of transformation and selection ( Figure 1 ) . In these cases it is possible that chromosomal mutations were responsible for rescuing the conditional lethal phenotypes , possibly by mutation of the Para promoter to mitigate repression . In three cases , no colonies were recovered at all ( Para-gyrA , metK , murA ) . For murA , this might have been due to poor transformation efficiency . Finally , four conditional lethal mutants ( Para-adk , dnaTC , proS , yeaZ ) were recovered repeatedly with a plasmid coding for the gene ybiV . We presumed that ybiV interfered with the function of the arabinose promoter: all five essential genes are functionally different , and expression of ybiV promoted growth of these mutants under all restrictive conditions . Furthermore , ybiV has been found previously in screens using a Para construct [25] . We thus excluded the HCS ybiV from the subsequent analyses . This left us with ten essential genes for which we had identified one or more HCS ( Figure 1 ) . For each of these ten genes , we subjectively chose one HCS for further analysis – with the exception of the essential gene degS , for which we included both HCS . Next , we tested whether the recovered HCS plasmids could replace the functions of their respective essential genes , or whether viability might rely on low-level transcription from the repressed Para promoter . We attempted to knock out the corresponding essential genes in strains harboring the HCS plasmid , with expression of the suppressor induced using 50 µM IPTG . We were successful in four cases ( these essential gene - HCS pairs were: dapA/nanA , spoT/mutT , pyrH/cmk and fldA/fldB; Figure 3 ) . We were not able to delete these four essential genes from a strain carrying an empty control plasmid . The other six essential genes could not be deleted from strains containing the HCS plasmids . This suggested that in the presence of the HCS , low-level expression of the essential genes was sufficient to allow growth . Without the HCS , this residual low-level expression did not allow growth ( Figure 4 ) . Alternatively , it is possible that the high copy suppressor increased expression from Para , and thus restored normal levels of the essential proteins . Therefore , we tested whether any of these seven HCS ( two HCS were included for degS ) increased expression from the Para promoter . We used a chromosomally encoded Para-phoA fusion to monitor expression from Para under conditions where expression of the HCS is induced . HCS clones overexpressing yciR , yhbJ , and ftnA exhibited slightly lower levels of PhoA-activity compared to controls , while overexpression of degP , rho and dpiA resulted in slightly elevated levels of PhoA-activity ( approximately 1 . 5-fold increase over the control ) . However , this activity was more than 50-fold below the activity of Para when induced with 0 . 1% L-arabinose ( Figure S1 ) . This suggested that none of these HCS rescued the conditional lethal phenotype through increasing expression of the essential gene . To test if the HCS genes caused a general non-specific rescue , we purified the HCS plasmids from the ASKA collection and transformed these plasmids into each conditional lethal mutant . We tested for regrowth in the same way as in the retransformation test described above . With one exception ( see below ) , none of the HCS plasmids restored growth in any other conditional lethal strain except the strain it was recovered from . Therefore we assumed that the observed high copy suppression is due to a specific mechanistic link between the depletion phenotype and the high copy suppressor ( Table 1 ) , rather than a consequence of a high copy suppressor-mediated induction of expression from Para . The single exception to this pattern was the HCS ftnA ( coding for ferritin ) , which rescued ftsK- or nrdAB depletion . ftnA and nrdB exhibit structural homology ( Table 2 ) , suggesting that this is a specific functional replacement . In the case of ftsK , the mechanism of suppression is less clear . One possibility is that the FtnA protein alleviates oxidative stress [26] that results from the loss of FtsK , as a consequence of double strand breaks in chromosomal DNA [27] . However , in both cases , the mechanism does not appear to be moderated through FtnA restoring expression from Para ( Figure S1 ) . These data thus show that out of the 23 essential genes or operons that we assessed , the functions of four could be completely replaced by non-orthologous genes . In six additional cases , the functions of the essential genes could be almost completely replaced: over-expression of a second gene enabled cellular viability even when the expression of the essential gene was largely abolished . In contrast , without overexpression of this second gene , no growth occurred . We also quantified how well the high copy suppressors restored growth in the four strains in which we could knock out the essential gene , by measuring how the growth yield depended on the dosage of HCS expression . The complemented knockout mutants showed qualitatively different responses to increasing expression of the HCS , as measured by the amount of the inducer IPTG added to the growth medium ( Figure 5 ) . The dapA knockout exhibits very low levels of growth at all levels of inducer , suggesting that NanA has only a low level of activity toward the DapA substrate , or that very high levels of activity are required to sustain growth . On the other hand , both the spoT and pyrH knockouts exhibit growth even when the suppressor is uninduced , suggesting that either these proteins are much more promiscuous , or that only low levels of activity are required for growth ( Figure 5 ) . To gain insight into the mechanisms of suppression , we compared the amino acid sequences ( using Smith-Waterman alignments [28] ) and protein structures ( using pairwise structural alignments [29]; Table S2 ) of all complementing HCS – essential gene pairs . Of the four gene pairs for which we could knock out the essential gene , two HCS share homology in amino acid sequence and protein structure with the suppressed essential gene: dapA and nanA share amino acid homology ( Table 2 ) , with nanA being the closest homologue of dapA in E . coli . However , phylogenetic analysis shows that they are only distantly related , and most likely diverged before the most recent common ancestor of all bacteria ( Figure S2B ) . The second HCS – essential gene pair showing amino acid homology is fldA and fldB ( Table 2 ) , with fldB being the closest homologue of fldA in the E . coli genome . Phylogenetic analysis suggests that these two proteins diverged after the origin of gamma-proteobacteria ( Figure S2A ) . In the set of six HCS for which we were unable to knock out the corresponding essential gene , one pair exhibits sequence homology: the proteins DegS and DegP [30] ( Table 2 ) . Again , degP ( with degQ ) is the closest homologue of degS in the E . coli genome . In several cases , the functions of the essential gene and its complementing HCS appear to be related . Besides having amino acid and structural similarities , the three pairs mentioned above ( dapA-nanA , fldA-fldB , and degS-degP ) have known functional similarities . DapA and NanA both belong to the N-acetylneuraminate lyase subfamily and catalyze similar biochemical reactions [31] , although dapA is essential and nanA is non-essential . It has previously been shown that a single amino acid exchange can turn NanA , an N-acetylneuraminate lyase , into an efficient dihydrodipicolinate synthase , which is the dedicated function of DapA . This amino acid exchange was hypothesized to optimize the turnover rate rather than the specificity of the reaction [32] . Therefore , replacement of DapA by NanA may be an example of how the increased concentration of an enzyme with promiscuous activity can promote an essential biochemical reaction and restore viability in the absence of the gene originally encoding the essential function . FldA and FldB are both flavodoxins . FldB is non-essential , in contrast to FldA [1] , and cannot replace the function of FldA when expressed from its native promoter [33] . However , our results show that that FldB can replace the function of FldA when expressed at a high level . DegS activates the sigma E stress response via proteolytic degradation of the anti-sigma factor RseA [34] triggered by misfolded outer membrane porins [35] , and it seems that DegP can fulfill the same function [30] . Although this implies that DegP overexpression might fully compensate for DegS function , we were unable to delete degS when overexpressing degP . Two of the essential gene – HCS pairs that exhibit no apparent amino acid or structural similarity do exhibit functional similarity: pyrH/cmk and spoT/mutT . cmk and pyrH both code for a nucleotide kinase: PyrH converts uracil monophosphate to uracil diphosphate , while Cmk converts cytosine monophosphate to cytosine diphosphate . It is also known that Cmk can use both cytidine and uridine ( the primary substrate of PyrH ) as substrates . Additionally , it has been shown previously that cmk can act as a high-copy suppressor of a temperature-sensitive pyrH allele [36] . Here we have shown that cmk is fully suppressive by deleting the entire pyrH locus . Both mutT and spoT can recognize phosphorylated guanosines as substrates , and cleave phosphoryl groups . SpoT is a key enzyme of the stringent response and hydrolyzes penta/tetra guanosine phosphate ( ( p ) ppGpp ) [37] , [38] . Deletion of spoT leads to the accumulation of ( p ) ppGpp , which in turn activates stringent response and leads to cessation of cell growth . One possible mechanism of suppression is that mutT can cleave phosphoryl groups from ( p ) ppGpp , converting it into a phosphorylated guanosine that no longer triggers the stringent response , thus allowing cell growth . Other possible functional similarities between pairs of high copy suppressors and essential genes are listed in Table 1 . We also objectively evaluated the hypothesis that there are specific functional relationships between essential genes and their HCS by testing whether the functional annotations of essential genes and their HCS are more similar than expected by chance . We collected all the GO molecular function annotations [39] for each essential gene and its corresponding HCS , and calculated a functional distance between each pair ( see Methods ) . Using this functional distance measure , we found that essential genes and their complementary HCS genes are much more similar in function than would be expected by chance ( p = 0 . 0024 , one-tailed Kolmogorov-Smirnov test ) . When we exclude the homologous pairs of genes ( degS-degP , dapA-nanA , and fldA-fldB , a weak signal of functional similarity remains ( p = 0 . 05 , one-tailed Kolmogorov-Smirnov test ) . Thus , it appears that the HCS genes compensate for the deleted essential genes through specific complementation of the missing function . On the other hand , ybiV , which is likely to be a non-specific HCS , shows no pattern of having greater functional similarity to its paired essential genes than would be expected by chance . Comparative genomic analyses have shown that genes that are essential in E . coli tend to be conserved in other bacterial taxa [12]–[15] . In addition , recent empirical results have shown that the essential functions of genes tend to be conserved: genes that are essential in one taxon have orthologues that are essential in other taxa ( Figure 1 ) . There are , however , exceptions: some essential genes are less well conserved , or have become non-essential in some taxa . Here , we have shown that under laboratory conditions , the functions of many essential genes can be completely or partially replaced by homologous or unrelated non-essential genes . A simple explanation might connect these two observations: if it is difficult to replace the function of an essential gene , then this gene should be both highly conserved and consistently essential across bacterial taxa . We thus asked whether it is more difficult to find genes providing compensatory functions for genes that are both conserved and consistently essential . We used data on conservation ( see Methods ) and empirical assessments of essentiality [1] , [3]–[11] to test this hypothesis . Of the 23 essential genes or operon pairs that we investigated , eight are both conserved and essential for all bacteria in which essentiality has been empirically assessed ( Figure 1 ) . Within this set , we found an HCS for only a single gene , ygjD . 15 of the essential genes or operons that we considered are either not fully conserved across bacteria or are nonessential in some taxa , or both . Of these , we found HCS for 9 of the 15; if we exclude the operon pairs , as it may generally be more difficult to find compensatory functions for both genes , this fraction increases to 8 out of 12 . The probability of finding so few HCS for conserved and consistently essential genes by chance is 0 . 037 and 0 . 025 , respectively ( one-tailed Fisher's exact test ) . The estimated odds ratios are ( with 95% upper limits in parentheses ) : 0 . 11 ( 0 . 89 ) and 0 . 083 ( 0 . 75 ) , respectively . These data suggest that suppressors of conserved consistently essential genes are approximately one fifth as likely to be found as suppressors for genes that are less conserved or are non-essential in some taxa . Thus , genes that are ancient , strongly conserved , and consistently essential across taxa appear to be persistently essential under laboratory conditions . The data here support the hypothesis that in some cases , simply increasing the expression level of specific non-essential genes can render essential genes non-essential ( i . e . high copy suppression ) . Sequence and structural comparisons showed that some of the HCS genes were homologous to the essential genes whose function they replaced . However , these homologues tended to be distantly related , with divergence times ranging from before the root of all bacteria , to soon after the origin of gamma-proteobacteria . In addition , in the majority of cases , homology was not required for HCS to occur , highlighting the possibility that even when there is no detectable homology , elevated expression can act as a mechanism allowing functional replacement . For example , although the essential gene – HCS pairs spoT-mutT and pyrH-cmk do not share detectable similarity on protein structure or sequence level , their biochemical activity is apparently similar enough to allow complete functional complementation . In several instances , the HCS that we recovered did not allow the deletion of the corresponding essential gene . We hypothesize that a combination of very low expression levels of the essential gene and expression of the relevant HCS allowed suppression of the lethal phenotype . However , in almost all cases , the suppression of the conditional lethal phenotype seemed to be based on a specific mechanistic link , a hypothesis that was further supported by the finding that the molecular functions of essential genes and their dedicated high-copy suppressors are far more similar than would be expected by chance . Finally , we have shown that when the function of an essential gene can be replaced in the laboratory , orthologues of that gene are more likely to be non-essential or absent from the genomes of other bacterial taxa . This observation suggests that compensability may influence patterns of evolutionary conservation: the functions of some essential genes are easier to replace than others , and the genes that perform such functions may be lost more often over evolutionary time . Previous studies have looked for high copy suppressors of lethal phenotypes [25] , [40] , [41] . The majority of these studies have been performed on a smaller scale or by screening for suppressors of mutations that cause non-lethal phenotypes , with the aim to investigate gene function . Our study is a comparative and systematic attempt to quantify the frequency of suppressors of essential genes , and to test if there is a statistical association between gene conservation and compensability . The potential for finding redundant , yet non-orthologous genes that can functionally replace essential genes might be a function of genome size . Previous work has shown that bacterial species with large genomes have fewer essential genes than species with small genomes [4] , [5] . One explanation for this observation is that in large genomes , there is a greater chance that a second gene encodes a similar function . Thus , the chance to replace essential gene functions with other functions could be greater in species with larger genome sizes and a generalist lifestyle . It would thus be interesting to test how the results of this study compare with additional studies in bacteria having much larger ( or smaller ) genomes . Indeed , in bacteria with small genomes , almost all essential genes are also highly conserved; thus finding conditionally essential genes may prove far more difficult . Overall , our work provides a novel explanation for the different patterns of conservation that are observed for essential genes , and emphasizes that gene essentiality is a fluid characteristic , even over short periods of evolutionary time .
All strains were grown in LB media ( Sigma ) or LB agar plates ( 1 . 5% agar , Sigma ) , and L-arabinose or D-glucose ( both Sigma ) was supplemented as indicated . E . coli strains MG1655 and DY330 were described previously [42] , [43] and grown at 37°C and 32°C , respectively , with vigorous shaking . AB330 is a Lac+ derivative of DY330 , and was received from Alex Boehm , University of Wurzburg , Germany . P1 transduction and TSS transformation were done as described elsewhere [44] , [45] . Strains harboring a pKD4 derived kanamycin resistance cassette were grown with 50 µg/ml kanamycin sulfate ( Sigma ) , and strains with ASKA ( - ) plasmids with 15 µg/ml chloramphenicol ( Calbiochem ) . Ampicillin ( Fluka ) 25 µg/ml was used to select for Para-phoA insertion in attB . Strains transformed with pCP20 [46] were grown at 32° in the presence of 15 µg/ml chloramphenicol ( Calbiochem ) . IPTG ( isopropyl thiogalactopyranoside ) was from Sigma . No comprehensive collection of conditional lethal mutants of essential genes is available . To construct a collection , we selected 23 essential genes and operon pairs from E . coli that exhibited varying levels of conservation across other bacterial taxa [15] . Genes were balanced for functional categories , but otherwise random . This group of essential genes covers nearly 10% of the essential gene content of E . coli MG1655 . Before we selected essential genes for our experiment , we discarded genes located in operons coding for other essential genes , because insertion of the Para construct in front or inside operons might have strong polar effects . Three exceptions were made: nrdAB , dnaTC and glmUS are essential tandem operons whose gene products interact physically or are involved in the same cellular processes . We assumed that the construct we use to repress transcription abolishes expression of both genes . We used the previously described strain TB55 [47] as PCR template for construction of arabinose-inducible conditional lethal mutants ( analogous to Roux et al . [48] ) , with the aim of tightly linking a kanamycin marker to the arabinose-inducible promoter of the araBAD operon . This strain allows the generation of a PCR product that contains an outward facing kanamycin resistance marker on one end , and on the other end an outward facing arabinose-inducible promoter . Insertion of this construct in front of essential genes and fusion of the Para promoter to transcriptional or translational start sites allows control of expression of selected essential genes [48] . We used TB55 to generate PCR products flanked by 40 to 42 base pairs homology to the upstream region of essential genes of interest . The PCR product spanned the kanamycin resistance gene , araC and the full intergenic region between araC and araB . Next , we constructed TB741 , a strain that allowed us to monitor expression of Para from a second , independent arabinose-inducible araBAD promoter . To that end , we combined a phoA knockout acquired from KEIO clone JW0374 [1] , and , after removal of the kanamycin resistance marker with pCP20 [46] , a Para-phoA construct was inserted into attB ( derived from E . coli strain SA22 ( a gift from Prof . Winfried Boos , University of Konstanz , Germany ) with P1 phage transduction . All strains used in this study can be found in Table S3 . All conditional lethal mutant strains were constructed initially with the same primer design , which included the following: deletion of 40 to 100 base pairs of the upstream region of the gene of interest by insertion of the PCR product generated from TB55 , and fusion of the start codon of the gene of interest with the start codon of araB . We were not able to recover clones with a conditional lethal character for yeaZ and murA following this methodology . Therefore we fused the transcription initiation site of araB to the predicted transcriptional start sites of murA and yeaZ ( from www . regulondb . ccg . unam . mx ) , yielding conditional lethal clones . All oligonucleotide sequences can be found in Tables S4 and S5 . As mentioned above , strain TB55 was used to generate PCR products that contained a kanamycin cassette adjacent to araC , the full Para-region and 42 to 45 base pairs at the 5′ and 3′ -prime ends that were homologous to the upstream and N-terminal region of the essential gene of interest . DY330 cells were grown in LB medium supplemented with 0 . 2% arabinose and made electro- and recombination competent as described previously [49] . After electroporation , cells were rescued in LB medium containing 0 . 2% arabinose and incubated at 32° for 1 . 5 hours prior to plating on arabinose- and kanamycin - containing LB plates . Clones were checked on LB plates supplemented with 0 . 4% glucose to confirm their conditional lethal character . The constructs were then moved by P1-transduction into TB741 , and conditional lethality was assessed again on LB plates with 0 . 4% glucose . All promoter fusions as well as the adjacent araC gene were verified by sequencing . We used the ASKA ( - ) strain collection [50] to construct a plasmid pool that contained all Escherichia coli open reading frames . The ASKA ( - ) library consists of 4123 clones , each one carrying a plasmid with one open reading frame . We pin-replicated clones into 96-well plates containing LB medium ( Sigma ) and 15 µg/ml chloramphenicol . Plates were incubated for 48 hours at 37°C . Then , 20 µl of each well were pooled , but clones containing plasmids that coded for essential genes of interest in our experiment were excluded . Plasmids were extracted using a plasmid preparation kit ( Promega ) , following the recommendations of the manufacturer . Each conditional lethal mutant was grown in LB medium with 0 . 1% arabinose ( Sigma ) to an OD600 nm of 0 . 4 to 0 . 8 . During the preparation of electrocompetent cells , the density of all cultures was adjusted to an OD600 nm of 1 to guarantee an equal number of cells per transformation event . Each clone was electroporated with 1 µl of plasmid pool ( DNA concentration approximately 330 ng/µl ) . Cells were rescued with 1 ml LB medium with 15 µg/ml chloramphenicol , and 100 µl was immediately removed and transferred to 900 µl LB medium with 0 . 1% arabinose and 15 µg/ml chloramphenicol to estimate transformation efficiency . To select for high copy suppressors , cells were spread on LB agar plates containing 0 . 4% glucose ( to enhance repression of Para ) , 50 µM IPTG and 15 µg/ml chloramphenicol , and incubated at 37°C until colonies appeared , or maximally 3 days to minimize the formation of colonies that might arise due to chromosomal suppressor mutations . We recovered up to 10 colonies per transformation event and restreaked them onto plates with glucose , IPTG , and chloramphenicol to verify growth . After successful regrowth , clones were grown in liquid cultured overnight with 0 . 1% arabinose and plasmid was extracted ( Promega ) . The purified plasmid was retransformed [44] into fresh ancestral conditional lethal mutant strains under permissive conditions , and 4 independent colonies of each transformation event were tested on permissive and restrictive plates for growth . This procedure directly tested for suppression mediated by more than one plasmid: only upon successful regrowth of all 4 clones , were the plasmids sequenced using the primer 5′-GCGGATAACAATTTCACACAGA-3′ . Cases in which all four clones did not grow were discarded from further analysis . The outcome of this retransformation test was based on the transformation method that exhibits a comparably low efficiency [44] , decreasing the probability of transforming two different plasmids into the same cell , making it unlikely that more than one plasmid was responsible for high copy suppression . After this verification procedure , we went back to the original ASKA ( - ) library , purified plasmids that we recovered from the screen ( except yciR; this gene was not contained in the clone at the indicated position in the collection ) , and repeated the procedure . This led to exclusion of two high copy suppressors for aspS , and verified all other suppressive plasmids we found in the screen . To determine if expression of a high copy suppressor lead to strongly increased expression of the Para promoter , we assayed the activity of a Para-phoA fusion inserted into the lambda attachment site , using a previously described procedure [51] . Briefly , we transformed plasmids coding for high copy suppressors ( and as a control the empty plasmid pCA24N ) into TB741 , and grew clones overnight with 50 µM IPTG , 0 . 4% glucose and 15 µg/ml chloramphenicol in 96 well plates , replicating each clone independently 16 times . To estimate the maximum expression level of Para-phoA , we induced 16 replicates of TB741 harboring pCA24N with 0 . 1% arabinose . After overnight growth , cultures were spun down , resuspended in phoA buffer ( 150 mM TrisHCl adjusted to pH 9 ) , diluted 1∶2 into fresh phoA buffer to a volume of 180 µl , and the OD600 nm was measured . One drop ( approximately 10 µl ) of a 1% SDS solution ( Sigma ) and 25 µl of a 10 mg/ml PNPP ( 4-nitrophenylphosphate , Sigma ) solution was added . After incubation at room temperature for 24 hours the OD550 nm and OD420 nm were measured , and PhoA activity determined using the formula described in [51] . To delete spoT , pyrH , fldA and dapA , plasmids encoding mutT , cmk , fldB and nanA were transformed into AB330 and expression was induced with 50 µM of IPTG ( or 1 mM for nanA ) . Knockouts were achieved following previously described methods [43] , [49] using a pKD4-derived kanamycin cassette flanked by homologous ends . Successful deletions were moved into MG1655 ( harboring ASKA ( - ) plasmids coding for HCS ) with P1 transduction [45] with addition of IPTG and verified by PCR using primers upstream and downstream of the insertion . To test for specific interactions between the depletion of essential genes and expression of plasmid-based non-complementing high copy suppressors , the HCS plasmids purified from the ASKA ( - ) library were transformed under permissive conditions into each conditional lethal mutant , and regrowth was checked as described for the initial screening procedure . Gene deletion mutants were grown overnight at 37°C in 96-well plates with shaking at 400 rpm in 8-fold replication , in LB medium supplemented with 1 mM IPTG and 15 µg/ml chloramphenicol . Cultures were spun down , washed once in LB medium , and diluted 1∶10−4 into fresh medium with IPTG concentrations as indicated . Optical density at 600 nm was measured every 30 minutes , for 7 . 5 hours in total . To analyze differential growth of conditional lethal mutant strains ( with suppressive plasmids , empty control plasmids , deletion of essential genes or ancestral conditional lethal mutants ) , we grew the corresponding clones in 96-well plates overnight . As a control , each conditional lethal mutant and the ancestral TB741 strain were transformed with the empty plasmid pCA24N . Conditional lethal mutants were grown with 0 . 1% L-arabinose and , if required , in presence of 15 µg/ml chloramphenicol to select for ASKA ( - ) plasmids . Essential gene deletion mutants were cultured with 1 mM IPTG to induce expression of high copy suppressors and to decrease the likelihood of genetic suppressor mutations . After overnight growth , cultures were serially diluted by repeatedly transferring 20 µl of culture into 180 µl of LB medium . Of this dilution series , 5 µl of the indicated dilutions were spotted onto plates supplemented with arabinose , glucose , chloramphenicol and IPTG as indicated and incubated as indicated . We used assignments based on previously published data [15] . Briefly , we used reciprocal shortest distance [52] to find potential orthologues of the relevant E . coli genes in the respective genomes . Two genes that are reciprocally the most closely related were denoted as orthologues if they aligned over more than 60% of the longer gene . In cases in which no orthologues were found , we used the MicrobesOnline database to search for genes named as putative orthologues . In this way , we found two additional putative orthologues , one for ftsK in S . pneumonia , and a second for plsC in S . aureus . We used data from ten empirical studies on essentiality [1] , [3]–[11] to determine whether or not genes orthologous to those in E . coli were essential in other bacterial taxa . Orthologous and homologous genes from a range of bacterial taxa were selected and aligned using Muscle v3 . 8 . 31 [53] with default parameters . The alignments were cleaned using GBlocks 0 . 91b [54] with length of non-conserved positions set to 32 , the number of flank and conserved positions set to minimum values , minimum block length to 2 , and allowed gaps set to all . This alignment was used as input into MrBayes 3 . 1 . 2 [55] with a mixed amino acid model and invariant plus gamma distributed rate variation across sites . The chains were run for 200 , 000 ( dapA/nanA ) , or 1 , 000 , 000 ( fldA/fldB ) generations , and the last 20% of the run was used for construction of a majority rule tree . We obtained molecular function annotations from the GO database ( www . geneontology . org/GO . downloads . annotations . shtml; 5/20/2011 ) for all annotated E . coli genes . We also obtained the relationships between all GO categories ( www . geneontology . org/GO . downloads . ontology . shtml; OBO v1 . 2 ) . GO annotations are related in a tree-like manner , beginning with broad , non-specific parent categories ( e . g . “binding” ) , each of which have more specific child categories ( e . g . “acyl binding” ) . Thus , we quantified functional distance as the number of parent categories that separate any two genes , normalized by the total number of parent categories for each gene . We calculated this distance between each essential gene and all other genes in the genome , and compared this to the distance between the essential genes and their complimentary HCS . This yields a number between 0 and 1 , specifying the fraction of genes in the genome that are less functionally similar than the essential gene and its HCS . If functional similarity does not play a role for the essential gene - HCS pairs , we would expect this number to be 0 . 5 , on average , and distributed uniformly between 0 and 1 . Instead , we found that for 12 out of 13 essential gene HCS pairs , this distance was less than 0 . 5 ( i . e . they were more similar than the average pair of genes ) ; for 8 out of 13 pairs , the distance was less than 0 . 25 . We used a Kolmogorov-Smirnov test to compare this distribution to the distribution expected if there were no functional relation between the essential gene and its HCS ( the uniform distribution ) . All statistical tests were done in R v2 . 13 . 1 [56] . | In any given organism , a fraction of all genes in the genome are required for viability; if they are experimentally deleted , the organism dies . Interestingly , the set of essential genes is usually not identical even for closely related organisms . Genes that are essential in one organism are sometimes nonessential in sister taxa or even missing from their genomes . This suggests that , in the course of evolution , some genes can be rendered non-essential and consequently can be lost . How can genes become non-essential ? It is possible that changes in an organism's living conditions render previously essential functions unessential . Alternatively , it is possible that , during evolution , the function of an essential gene can be taken over by another gene , so that the essential gene becomes dispensable . Here , we tested the second hypothesis experimentally in the laboratory . We tried to replace the functions of essential genes in the bacterium Escherichia coli . We find that the genes that can easily be replaced in the laboratory are also more likely to be lost in the course of evolution . This suggests that differences in the evolutionary fate between essential genes can be partially explained by how easily their functions can be taken over by other genes . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"biology",
"microbiology",
"evolutionary",
"biology"
] | 2012 | Patterns of Evolutionary Conservation of Essential Genes Correlate with Their Compensability |
Local excitatory connections in mouse primary visual cortex ( V1 ) are stronger and more prevalent between neurons that share similar functional response features . However , the details of how functional rules for local connectivity shape neuronal responses in V1 remain unknown . We hypothesised that complex responses to visual stimuli may arise as a consequence of rules for selective excitatory connectivity within the local network in the superficial layers of mouse V1 . In mouse V1 many neurons respond to overlapping grating stimuli ( plaid stimuli ) with highly selective and facilitatory responses , which are not simply predicted by responses to single gratings presented alone . This complexity is surprising , since excitatory neurons in V1 are considered to be mainly tuned to single preferred orientations . Here we examined the consequences for visual processing of two alternative connectivity schemes: in the first case , local connections are aligned with visual properties inherited from feedforward input ( a ‘like-to-like’ scheme specifically connecting neurons that share similar preferred orientations ) ; in the second case , local connections group neurons into excitatory subnetworks that combine and amplify multiple feedforward visual properties ( a ‘feature binding’ scheme ) . By comparing predictions from large scale computational models with in vivo recordings of visual representations in mouse V1 , we found that responses to plaid stimuli were best explained by assuming feature binding connectivity . Unlike under the like-to-like scheme , selective amplification within feature-binding excitatory subnetworks replicated experimentally observed facilitatory responses to plaid stimuli; explained selective plaid responses not predicted by grating selectivity; and was consistent with broad anatomical selectivity observed in mouse V1 . Our results show that visual feature binding can occur through local recurrent mechanisms without requiring feedforward convergence , and that such a mechanism is consistent with visual responses and cortical anatomy in mouse V1 .
Much of our current understanding of local cortical connectivity in neuronal circuits of the neocortex is based on the presumption of randomness . Anatomical methods for estimating connection probabilities [1 , 2] and techniques for using anatomical reconstructions to build models of cortical circuits [3–7] are largely based on the assumption that connections between nearby neurons are made stochastically in proportion to the overlap between axonal and dendritic arborisations [8] . On the other hand , a wealth of evidence spanning many cortical areas and several species indicates that cortical connectivity is not entirely random . In species that display smooth functional maps in primary visual cortex ( V1 ) , such as cat and macaque monkey , long-range intrinsic excitatory connections tend to preferentially connect regions of similar function [9–13] . Although rodents exhibit a mapless , “salt and pepper” representation of basic visual features across V1 [14] , non-random connectivity is nonetheless prevalent both within and between cortical layers [15–20] , reflecting similarities in functional properties [21–25] or projection targets [26–28] . Despite multiple descriptions of specific connectivity in cortex , the rules underlying the configuration of these connections are not entirely clear . Whereas strong connections are more prevalent between neurons with similar receptive fields , the majority of synaptic connections are made between neurons with poorly-correlated receptive fields and poorly correlated responses [24] . This sea of weak synaptic inputs might be responsible for non-feature-specific depolarisation [24] or might permit plasticity of network function [20] . However , another possibility is that weak local recurrent connections reflect higher-order connectivity rules that have not yet been described . Recent reports have highlighted the facilitatory and selective nature of plaid responses in mouse V1 [29–31] . Many neurons in mouse V1 respond to plaid stimuli in accordance with a simple superimposition of their responses to the two underlying grating components ( i . e . “component cell” responses [32] ) . However , a significant proportion of neurons that are visually responsive , reliable and selective exhibit complex responses to plaid stimuli that are difficult to explain with respect to simple combinations of grating components [30] . We hypothesised that responses to complex stimuli in mouse V1 could be a result of local combinations of visual features , through structured local recurrent excitatory connectivity . These rules could be difficult to detect through anatomical measurements , if they comprised only small deviations from predominantly like-to-like connectivity . Here we examined whether small tweaks to recurrent connectivity rules could alter visual representations in cortex , by analysing the computational properties of cortical networks with defined rules for local connectivity . We simulated visual responses to grating and plaid stimuli in large networks with properties designed to resemble the superficial layers of mouse V1 , assuming distinct connectivity schemes . We then compared the response patterns and visual representations predicted by the network simulations with those recorded in vivo in mouse V1 , to test the predictions arising from our models . Specifically , we evaluated two broad classes of connectivity patterns , where specific local excitatory connectivity is defined according to the visual response properties of neurons ( Fig 1 ) : Despite the small difference in network configuration , these distinct rules give rise to radically different visual representations of plaid stimuli , both in terms of complexity of visual response selectivity of individual neurons and regarding facilitation versus suppression in response to these compound stimuli . We found that the complexity of plaid responses in mouse V1 was reproduced in our simulations when assuming the feature-binding connectivity scheme , with local connections grouping multiple feedforward response properties , but not when assuming purely like-to-like connections .
Under the assumption that the configuration of local recurrent connections in cortex might lead to differential processing of simple and compound visual stimuli , it is important to quantify the relationship between responses to grating and plaid stimuli in visual cortex . Plaid stimuli are often constructed from a single choice of relative component angle ( 90° orthogonal gratings ) , leaving open the possibility that a richer set of plaid stimuli would help to classify neurons with complex responses . We therefore probed mouse V1 with grating component stimuli composed of grating stimuli with 16 drift directions , and three full sets of plaid stimuli composed of 45° , 90° and 135° relative grating component orientations . We recorded responses from layer 2/3 neurons in V1 using two-photon imaging of animals expressing GCaMP6m ( Fig 2A–2F; 8 animals , 8 sessions , 441 / 879 responsive / imaged neurons; see Methods ) . We defined a modulation index ( MI ) to quantify the degree of facilitation or suppression elicited by plaid stimuli over grating stimuli , for single cortical neurons; large positive values for MI indicate strong facilitation in response to plaid stimuli , whereas large negative values indicate strong suppression ( see Methods ) . Visual responses to the full set of plaid stimuli were dominated by facilitation , and were significantly more facilitatory than when considering only the set of 90° plaids ( Fig 2I; median modulation index MI 0 . 098 ± [0 . 081 0 . 12] vs 0 . 011 ± [-0 . 0060 0 . 027]; p < 1⨉10–10 , Wilcoxon rank-sum; all following values are reported as median ± 95% bootstrap confidence intervals unless stated otherwise ) . The presence of stronger facilitation when comparing responses to the full set of plaid stimuli with responses to 90° plaids alone , is consistent with our earlier finding that some neurons in mouse V1 are highly selective for particular combinations of grating components [30] . Accordingly , we used a plaid selectivity index ( PSI ) to quantify how selective were the responses of single neurons over the set of plaid stimuli ( see Methods ) . The PSI was defined in analogy to orientation or direction selectivity indices ( OSI or DSI ) , such that values of PSI close to 1 indicate that a neuron responds to only a single plaid stimulus out of the set of presented plaid stimuli . Values of PSI close to 0 indicate that a neuron responds equally to all plaid stimuli . Responses to the full set of plaid stimuli were highly selective; significantly more selective than predicted by a component model generated using all plaid and grating stimuli ( Fig 2J; median PSI 0 . 38 ± [0 . 36 0 . 41] vs 0 . 30 ± [0 . 28 0 . 31]; p < 1⨉10–10 , Wilcoxon rank-sum ) and indeed significantly more selective than responses to the 90° plaids alone ( Fig 2J; median 90° PSI 0 . 25 ± [0 . 23 0 . 28]; p < 1⨉10–10 vs all plaids , Wilcoxon rank-sum ) . Therefore , probing visual cortex with a dense set of plaid stimuli reveals richer visual responses than when probed with a set of only 90° plaids . Indeed , recent results suggest that using an expanded set of plaid stimuli evokes more pattern-cell responses in mouse V1 [31] . Consistent with this finding , our results show that using a dense set of plaids does not make responses to compound stimuli trivial to predict based on component responses . In addition , we found that visual responses were more facilitatory and more selective than when measured using 90° plaids alone . How are selective , facilitatory responses to plaid stimuli generated in V1 ? As we suggested previously , one possibility is that specific grating component representations are combined through local excitatory connectivity [30] . In mouse V1 , synaptic connection probability is enhanced by similarity of orientation preference [21 , 23 , 25] , suggesting that local excitatory connections may group together neurons with common preferred orientations . Connection probability is even more strongly modulated by neuronal response correlations to natural visual stimuli; i . e . , the likelihood for a synaptic connection is higher for neuronal pairs responding similarly to natural scenes [21 , 22 , 24] . If connections in mouse V1 were strictly governed by preferred orientation , then neurons with similar orientation preference should also predominately have similar responses to natural movies , and vice versa . We recorded visual responses in populations of neurons labelled with the synthetic calcium indicator OGB in anesthetized mouse V1 ( 5 animals , 129 / 391 responsive neurons with overlapping receptive fields / total imaged neurons; S1A–S1C Fig; see Methods ) . We used signal correlations to measure the similarity between the responses of pairs of neurons with identified receptive fields ( S1A Fig ) to drifting grating ( S1B Fig ) and natural movie ( S1C Fig ) visual stimuli ( see Methods ) . We found that neuronal pairs with high signal correlations to natural scenes , which are most likely to be connected in cortex [21 , 22 , 24] , showed only a weak tendency to share similar orientation preferences ( S1D and S1E Fig; pairs with OSI > 0 . 3; p = 0 . 8 , Kruskall-Wallis ) . This is consistent with earlier findings in cat area 17 ( V1 ) , which showed a poor relationship between responses to gratings and natural movies [34] . Similarly , under a like-to-like connectivity rule , synaptically connected neurons in mouse V1 should share both similar orientation preference and responsiveness to natural movies . We therefore compared response correlations and preferred orientations for pairs of mouse V1 neurons , which were known to be connected from in vivo / in vitro characterisation of functional properties and connectivity ( data from [24] used with permission; 17 animals , 203 patched and imaged cells , 75 connected pairs ) . Consistent with our results comparing responses to gratings and natural movies , connected pairs of cells with similar orientation preference were not more likely to share a high signal correlation to flashed natural scenes ( S1F Fig; p = 0 . 54 , Kruskall-Wallis ) . Also consistent with earlier findings [21 , 23] , we observed a positive relationship between synaptic connectivity and similarity of orientation preference ( S1G Fig; p = 0 . 045 , Ansari-Bradley test ) . However , strongly connected pairs ( strongest 50% of excitatory post-synaptic potentials—EPSPs—over connected pairs ) were not more similar in their preferred orientation than the remaining pairs ( p = 0 . 17 , Ansari-Bradley test vs weakest 50% of connected pairs ) . Connected pairs spanned a wide bandwidth of preferred orientations , with more than 20% of connections formed between neurons with orthogonal preferred orientations . Spatial correlation of receptive fields is a comparatively better predictor for synaptic connectivity than shared orientation preference , but a majority of synaptic inputs are nevertheless formed between neurons with poorly- or un-correlated responses [24] . We conclude that similarity in orientation preference only partially determines connection probability and strength between pairs of neurons in mouse V1 . This weak functional specificity for similar visual properties can be explained by two possible alternative connectivity rules . In the first scenario , local excitatory connections in cortex are aligned with feedforward visual properties , but with broad tuning ( Fig 1A; a like-to-like rule ) . As a consequence , all connections show an identical weak bias to be formed between neurons within similar tuning , and the average functional specificity reported in S1G Fig and elsewhere [21 , 24] reflects the true connection rules between any pair of neurons in cortex . Alternatively , local excitatory connections may be highly selective , but follow rules that are not well described by pairwise similarity in feedforward visual properties . For example , subpopulations of connected excitatory neurons might share a small set of feedforward visual properties , as opposed to only a single feedforward property ( Fig 1B; a feature-binding rule ) . In this case , connections within a subpopulation could still be highly specific , but this specificity would be difficult to detect through purely pairwise measurements . If pairwise measurements were averaged across a large population , any specific tuning shared within groups of neurons would be averaged away . We designed a non-spiking model of the superficial layers of mouse V1 , to explore the effect of different connectivity rules on information processing and visual feature representation within the cortex . Non-spiking linear-threshold neuron models provide a good approximation to the input current to firing rate ( I–F ) curves of adapted cortical neurons [35]; model neurons with linear-threshold dynamics can be directly translated into integrate-and-fire models with more complex dynamics [36 , 37] , and in addition form good approximations to conductance-based neuron models [38] . A full list of parameters for all models presented in this paper is given in Table 1 . The dynamics of neuronal networks defined with particular specific synaptic connectivity rules remain generally unknown , although some results suggest that specific connectivity leads to reduced dimensionality of network activity patterns [40] . Here we explored the relationship between specific connectivity and network dynamical properties in a non-linear , rate-based network model incorporating realistic estimates for recurrent excitatory and inhibitory connection strength in layer 2 / 3 of mouse V1 . To explore the basic stability and computational consequences of functionally specific excitatory connectivity , we built a small five-node model ( four excitatory and one inhibitory neurons; “analytical model”; Fig 3 ) . Connections within this model were defined to approximate the average expected connectivity between populations of neurons in layer 2 / 3 of mouse V1 . Excitatory neurons were grouped into two subnetworks , and a proportion s of synapses from each excitatory neuron was reserved to be made within the same subnetwork . When s = 0 , E↔E synapses were considered to be made without specificity , such that each connection in the small model approximated the average total connection strength expected in mouse V1 in the absence of functional specificity . When s = 1 , all E↔E synapses were considered to be selectively made within the same subnetwork , such that no synapses were made between excitatory neurons in different subnetworks . Connections to and from the inhibitory node were considered to be made without functional specificity in every case , mimicking dense inhibitory connectivity in mouse visual cortex [41–44] . The general form of the weight matrix is therefore given by W=[aabb−wieaabb−wiebbaa−wiebbaa−wieweiweiweiwei−wI∙fI] ( 2 ) where wS = wE ⋅ ( 1− fI ) ⋅s is the specific weight component , wN = wE ⋅ ( 1− fI ) ⋅ ( 1−s ) is the nonspecific weight component , wE is the total synaptic weight from a single excitatory neuron , wI is the total synaptic weight from a single inhibitory neuron; fI = 1/5 is the proportion of inhibitory neurons; a = wS / 2+wN / 4 is the excitatory weight between neurons in the same subnetwork; b = wN / 4 is the excitatory weight between neurons in different subnetworks; wie = wI ⋅ ( 1− fI ) / 4 is the nonspecific inhibitory to excitatory feedback weight; and wei = wE ⋅ fI is the nonspecific excitatory to inhibitory weight . Amplification in the network with specific connectivity is selective ( Fig 3B and 3C ) : neurons within a subnetwork recurrently support each other’s activity , while neurons in different subnetworks compete . Therefore , which sets of neurons will be amplified or will compete during visual processing will depend strongly on the precise rules used to group neurons into subnetworks . We therefore examined the impact of like-to-like and feature-binding rules on responses in our analytical model . The excitatory network was partitioned into two subnetworks; connections within a subnetwork corresponded to selective local excitatory connectivity within rodent V1 . Under the like-to-like rule , neurons with similar orientation preferences were grouped into subnetworks ( Fig 3D ) . We tested the response of this network architecture to simulated grating and plaid stimuli , by injecting currents into neurons according to the similarity between the orientation preference of each neuron and the orientation content of a stimulus ( see grating labels in Fig 3D and 3E ) . When a stimulus matched the preferred orientation of a neuron , a constant input current was injected ( Ii ( t = ι ) ; when a stimulus did not match the preferred orientation , no input current was provided to that neuron ( Ii ( t ) = 0 ) . When simulating the analytical model , the input current ι = 1 . Under the like-to-like rule , responses of pairs of neurons to simple grating stimuli and more complex plaid stimuli were highly similar ( Fig 3D ) . Amplification occurred within subnetworks of neurons with the same preferred orientation , and competition between subnetworks with differing preferred orientation [53 , 55] ( visible by complete suppression of response of neurons in lower traces of Fig 3D ) . Alternatively , we configured the network such that the rules for local excitatory connectivity did not align with feedforward visual properties ( a feature-binding rule ) . We formed subnetworks by grouping neurons showing preference for either of two specific orientations ( Fig 3E ) . When this feature-binding connectivity rule was applied , neuronal responses to grating and plaid stimuli differed markedly ( cf . top vs bottom panels of Fig 3E ) . Selective amplification was now arrayed within populations of neurons spanning differing orientation preferences , and competition occurred between subnetworks with different compound feature preferences . Importantly , a feature-binding rule implies that neurons with the same preferred orientation could exist in competing subnetworks . While their responses to a simple grating of the preferred orientation would be similar and correlated ( Fig 3E; indicated by a high response correlation measured over grating responses ρg ) , the same two neurons would show decorrelated responses to a plaid stimulus ( Fig 3E; indicated by a low response correlation measured over plaid responses ρp ) . We conclude that changes in pairwise response similarity , provoked by varying the inputs to a network , can provide information about the connectivity rules present in the network . The results of our simulations of small networks suggest that rules for specific local connectivity can modify the correlation of activity between two neurons in a network , depending on the input to the network . The question arises of how connectivity rules shape distributed representations of visual stimuli , when examined across a large network and over a broad set of stimuli . We therefore simulated the presentation of grating and plaid visual stimuli in a large-scale non-linear , rate-based model of the superficial layers of mouse V1 . Individual neurons were modelled as described above for the small scale network ( Eq ( 1 ) ) . To construct the large-scale simulation model of mouse V1 , 80 , 000 linear-threshold neurons were each assigned a random location in physical space ui∈T2 where T defines the surface of a virtual torus of size 2 . 2×2 . 2 mm . Excitatory and inhibitory neurons were placed with relative densities appropriate for layers 2 and 3 of mouse cortex [56] . Approximately 18% of neurons were inhibitory; [57 , 58]; see Tables 1 and 2 for all parameters used in these models . Excitatory neurons were assigned an orientation preference θ drawn from a uniform random distribution , mimicking the “salt and pepper” functional architecture present in rodent visual cortex [14] . We simulated the presentation of grating and plaid stimuli in our large-scale network model of mouse V1 . We quantified response similarity between pairs of neurons as suggested by the results of the small network simulations: by measuring pairwise response correlations over a set of grating stimuli ( ρg ) , and separately over a set of plaid stimuli ( ρp; see Methods ) . In the network that implemented a like-to-like connection rule for recurrent excitatory connectivity ( Fig 4A and 4B ) , pairs of neurons showed similar responses to both grating and plaid stimuli ( Fig 4B; R2 = 0 . 83 between ρg and ρp ) , in agreement with the analytical like-to-like model of Fig 3D . However , in the network that implemented a feature-binding connection rule , where in addition to spatial proximity and similarity in preferred orientation subnetworks were defined to group neurons of two distinct preferred orientations ( Fig 4C and 4D ) , neurons showed reduced correlation in response to plaid stimuli ( Fig 4D , R2 = 0 . 13 between ρg and ρp ) , in agreement with the analytical feature- binding model of Fig 3E . Different configurations of local recurrent excitatory connectivity produced by like-to-like or feature-binding rules can therefore be detected in large networks , by comparing responses to simple and compound stimuli . Consistent with our analytical models , networks without functionally specific connectivity did not give rise to decorrelation ( S3B Fig; R2 = 0 . 72 between ρg and ρp ) . This shows that decorrelation between plaid and grating responses in our models does not arise simply due to random connectivity , but requires the active mechanism of selective amplification through feature-binding subnetwork connectivity . Inhibitory responses were untuned in our simulations ( blue traces in Fig 4A and 4C ) , in agreement with experimental observations of poorly-tuned inhibition in mouse V1 [42 , 58 , 65 , 66] . Our analytical network results show that in principle the configuration of local excitatory connectivity , whether aligned with or spanning across feedforward visual properties , has a strong effect on visual representations ( Fig 3 ) . Our large-scale simulations show that these effects can be detected in large networks as differences in the pairwise correlations of responses to simple and compound visual stimuli ( Fig 4 ) . We therefore aimed to test which connectivity scheme is more likely to be present in visual cortex , by examining responses of neurons in mouse V1 . Using two-photon calcium imaging , we recorded responses of populations of OGB-labelled neurons in mouse V1 to a set of contrast-oscillating oriented grating stimuli over a range of orientations , as well as the responses to the set of plaid stimuli composed of every possible pair-wise combination of the oriented grating stimuli ( Fig 5; 5 animals , 5 sessions , 313 / 543 responsive / total imaged neurons; see Methods ) . Responses to plaid stimuli in mouse V1 suggest that stimulating with a denser sampling of compound stimulus space leads to a better characterisation of response selectivity [31] ( Fig 2 ) . Accordingly , we probed responses in mouse V1 under stimuli analogous to those used in the model simulations , with a dense coverage of plaid combinations over a set of finely-varying grating orientations . We found that consistent with our earlier findings examining 90° drifting plaid stimuli [30] , responses to grating stimuli did not well predict responses to plaid stimuli . Pairs of neurons with similar preferred orientation but with highly differing responses to plaid stimuli were common ( Fig 5B and 5C; R2 = 0 . 05 between ρg and ρp; OSI > 0 . 3 ) . The degree of decorrelation we observed in mouse V1 was considerably higher than predicted by the like-to-like model , and was more consistent with the feature-binding model ( Fig 5E ) . Decorrelation induced by plaid responses and the lack of a relationship between grating and plaid responses in mouse V1 were not a result of unreliable or noisy responses in vivo . We included in our analysis only neurons that were highly reliable , and responded significantly more strongly than the surrounding neuropil ( see Methods ) . As a further control , we used experimentally recorded responses to grating stimuli to generate synthetic plaid responses for mouse V1 that would result from a cortex with like-to-like subnetwork connectivity ( Fig 5D , inset; see Methods ) . Our control data were generated from single-trial responses of single V1 neurons , and therefore included the same trial-to-trial variability exhibited by cortex . This control analysis indicates that a like-to-like rule among V1 neurons would result in a higher correlation of grating and plaid responses than experimentally observed ( Fig 5D; median R2 = 0 . 77 ± [0 . 767 0 . 775] between ρg and ρp; n = 2000 bootstrap samples; compared with R2 = 0 . 05 for experimental results; p < 0 . 005 , Monte-Carlo test ) . Importantly , this control analysis is not restricted to our like-to-like rule , but makes similar predictions of highly correlated grating and plaid responses for any arbitrary model that combines grating components to produce a plaid response , as long as that rule is identical for every neuron in the network [30] . This is because if a single consistently-applied rule exists , then any pair of neurons with similar grating responses ( high ρg ) will also exhibit similar plaid responses ( high ρp ) . In contrast , neurons that are connected within the feature-binding model combine different sets of grating components , depending on which subnetwork the neurons are members of . Neurons in mouse V1 exhibited a wide range of facilitatory and suppressive responses to plaid stimuli , roughly equally split between facilitation and suppression ( Fig 5F and 5G; 45% vs 42%; MI > 0 . 05 and MI < –0 . 05 ) . The proportion of facilitating and suppressing neurons in mouse V1 was similar to that exhibited by responsive neurons in our feature-binding model ( Fig 5G; V1 versus F . B . , p = 0 . 17; two-tailed Fisher’s exact test , nV1 = 313 , nF . B . = 809 ) . In contrast , neither the like-to-like model nor a model without functionally specific connectivity exhibited significant facilitation in responsive neurons , and both were significantly different from the distribution of facilitation and suppression in mouse V1 ( Fig 5G; p < 0 . 001 in both cases; two-tailed Fisher’s exact test , nL-to-L = 729 , nRnd = 729 ) . The wide range of facilitatory and suppressive responses observed in mouse V1 is more consistent with a feature-binding rule for local connectivity , compared with a like-to-like rule or a network without functionally specific connectivity .
We found that the precise rules that determine local connections among neurons in cortex can strongly affect the representation of visual stimuli . The feature-binding rule we examined embodies the simplest second-order relationship between connectivity and preferred orientation , and was chosen for this reason . We cannot rule out more complicated connectivity rules as being present in mouse V1 , but we have shown that a simple like-to-like rule cannot explain responses to plaid visual stimuli . Random , non-functionally specific connections were also unable to explain complex plaid responses in mouse V1 ( S3 Fig ) . How can the detailed statistics of “feature-binding” rules be measured in cortex ? Existing experimental techniques have been used to measure only first-order statistical relationships between function and cortical connectivity [18 , 21–24 , 42] . Unfortunately , current technical limitations make it difficult to measure more complex statistical structures such as present under a feature-binding connectivity rule . Simultaneous whole-cell recordings are typically possible from only small numbers of neurons , thus sparsely testing connectivity within a small cohort . Even if simultaneous recordings of up to 12 neurons are used [17] , identifying and quantifying higher-order statistics in the local connectivity pattern is limited by the low probability of finding connected excitatory neurons in cortex . Nevertheless , our feature-binding connectivity model is consistent with the results of functional connectivity studies ( S1 Fig ) . In addition , our results highlight that small changes in the statistics of local connectivity can have drastic effects on computation and visual coding . Introducing a small degree of specificity , such that a minority of synapses are made within an excitatory subnetwork , is sufficient to induce strong specific amplification and strong competition to the network , even though a majority of the synapses are made randomly without functional specificity ( Fig 3A–3C ) . Under our feature-binding model 68% of synapses are made randomly; approximately 27% are made under a like-to-like rule and the remaining 5% are used to bind visual features . Clearly , detecting the small proportion of synapses required to implement feature binding in V1 will be difficult , using anatomical sampling techniques that examine only small cohorts of connected neurons . A recent study functionally characterised the presynaptic inputs to single superficial-layer neurons in mouse V1 , using a novel pre-synaptic labelling technique [67] . Consistent with our results for preferred orientation ( S1F and S1G Fig ) , they found that presynaptic inputs were similarly tuned as target neurons but over a wide bandwidth . The majority of synaptically connected networks were tuned for multiple orientation preferences across cortical layers , similar to the feature-binding networks in our study . We implemented an alternative approach , by inferring the presence of higher-order connectivity statistics from population responses in cortex . This technique could be expanded experimentally , by presenting a parameterised battery of simple and complex stimuli . Stimuli close to the configuration of local connectivity rules would lead to maximal facilitation and competition within the cortical network . Importantly , our results strongly suggest that simple stimuli alone are insufficient to accurately characterise neuronal response properties in visual cortex . Our theoretical analysis and simulation results demonstrate that functionally specific excitatory connectivity affects the computational properties of a cortical network by introducing amplification of responses within subnetworks of excitatory neurons , and competition in responses between subnetworks ( Fig 3A–3C ) . Several recent studies have demonstrated that visual input is amplified within the superficial layers of cortex [68–70] , and recent results from motor cortex suggest competition between ensembles of neurons [71] . Our modelling results indicate that some form of selective local excitatory connectivity is required for such amplification to occur through recurrent network interactions , under reasonable assumptions for anatomical and physiological parameters for rodent cortex ( Fig 3A–3C; S2 Fig ) . This still leaves in question whether the particular configuration of selective excitatory connectivity plays a role . Our simulation results showed that the effects of amplification and competition on cortical responses are tuned to the statistics of local connectivity . This implies that complex visual stimuli for which the composition of stimulus components matches the statistics of a subnetwork will undergo stronger amplification than other non-matching visual stimuli ( Fig 6 ) . In our feature-binding model , the statistics of subnetwork connectivity were defined to reflect combinations of two preferred orientations chosen from a uniform random distribution . This combination of two orientations is similar to the visual statistics of plaid stimuli with arbitrarily chosen grating components . As a result , plaid stimuli gave rise to stronger amplification than single grating components alone , when the composition of the plaid matched the composition of connectivity within a particular subnetwork . This led to a facilitatory effect , where some neurons responded more strongly to plaid stimuli than to the grating components underlying the plaid stimuli . Conversely , competition between subnetworks led to weaker responses to some plaid stimuli , for neurons that “lost” the competition . Competition could therefore be one cortical mechanism underlying cross-orientation suppression in response to plaid stimulation . In contrast , suppression in the like-to-like and random non-specific models occur because the energy in the stimulus is spread across two grating components , and is not combined by the network to form strong plaid selectivity . In the like-to-like model , competition occurs between representations of the two oriented grating components of the plaid , causing additional suppression . The presence of amplified , strongly facilitating plaid responses in mouse V1 is therefore consistent with the existence of subnetworks representing the conjunction of differently-oriented edges . Could the complexity of plaid texture responses in mouse V1 be explained by convergence of differently tuned feedforward inputs from layer 4 onto single layer 2/3 neurons , similar to the proposed generation of pattern-selective responses in primate MT [32 , 72] ? Building plaid responses in this way would imply that layer 2/3 neurons would respond to multiple grating orientations , since they would receive approximately equal inputs from at least two oriented components . However , layer 4 and layer 2/3 neurons are similarly tuned to orientation in rodent V1 [63 , 64] , in conflict with this feedforward hypothesis . In addition , if responses to complex stimuli were built by feedforward combination of simple grating components , then the response of a neuron to the set of grating stimuli would directly predict the plaid response of that neuron . This would then imply that two neurons with similar responses to plaid stimuli must have similar responses to grating stimuli . However we found this not to be the case experimentally; two neurons with similar responses to grating components often respond differently to plaid textures or to natural scenes ( S1D Fig; Fig 5A and 5B; [30] ) . We cannot rule out the influence of feedback projections on shaping responses to plaid stimuli . The time resolution of calcium imaging is too slow to differentiate between feedforward , recurrent local , and feedback responses based only on timing . However , top-down feedback inputs are considered to be suppressed during anesthesia [73]; in contrast , we observed complex responses to plaid stimuli in anesthetized animals . Since our proposed mechanism for feature binding relies on recurrent amplification , relatively few excitatory synapses are required to reproduce complex plaid responses . In contrast , non recurrent influences such as feedforward or feedback projections would require comparatively more synapses to achieve a similar pattern of plaid responses . There are more local recurrent excitatory synapses in V1 layer 2 / 3 than there are available excitatory synapses in feedback projections to V1 ( 22% recurrent excitatory synapses in layer 2 / 3 vs a maximum of 17 . 2% feedback synapses; [1] ) . In addition , putative feedback inputs would need to be wired with high functional specificity; this degree of anatomical specificity has not been demonstrated experimentally . Non-specific connectivity between excitatory and inhibitory neurons , as assumed in our simulation models , is consistent with the concept that inhibitory neurons simply integrate neuronal responses in the surrounding population [74] , and is also consistent with experimental observations of weakly tuned or untuned inhibition in rodent visual cortex [42 , 52 , 58 , 65 , 66] . Although specific E↔I connectivity has been observed in rodent cortex [16 , 28] , the majority of E↔I synapses are likely to be made functionally non-specifically in line with the high convergence of E→I and I→E connections observed in cortex [41 , 42 , 65] . In our models , shared inhibition is crucial to mediate competition between excitatory subnetworks ( Fig 3 ) ; inhibition is untuned because excitatory inputs to the inhibitory population are pooled across subnetworks . Poorly tuned inhibition , as expressed by the dominant class of cortical inhibitory neurons ( parvalbumin expressing neurons ) , therefore plays an important computational role and is not merely a stabilising force in cortex . Other inhibitory neuron classes in cortex ( e . g . somatostatin or vaso-intestinal peptide expressing neurons ) have been shown to exhibit feature-selective responses [58 , 75 , 76] . Recent computational work examined the influence of multiple inhibitory neuron classes with different physiological and anatomical tuning properties in a model for rodent cortex [77] . They examined the role of inhibitory connectivity on divisive and subtractive normalisation of network activity in a network with specific , orientation-tuned inhibitory connectivity . They found that specific inhibitory feedback could lead to divisive normalisation of network activity , while non-specific inhibitory feedback could lead to subtractive normalisation . However , the computational role of specific inhibition is likely to rest on the precise rules for connectivity expressed between excitatory and inhibitory neurons . If the rules for E↔E and E↔I connections align , then a specific inhibitory population could act as a break on excitation within a subnetwork , and could allow more specific anatomical connectivity to persist while maintaining the balance between excitation and inhibition in cortex . The functional profile of this balancing pool would be highly tuned , and be similar to that of the excitatory neurons in the subnetwork , suggesting a physiological signature of specific inhibitory feedback that could be sought experimentally . Alternatively , if E↔I connection rules result in counter-tuned specificity , these connections would act to strengthen competition between subnetworks . As discussed above , our like-to-like model of orientation-tuned selective excitatory connectivity coupled with non-specific inhibitory feedback is similar in network topology to classical ring models of orientation tuning in visual cortex ( e . g [53 , 54 , 78] ) . The principal difference in our model is the embedding of functionally selective connectivity within spatially-constrained anatomical connectivity . We showed that under model parameters chosen to be realistic in mouse V1 , only a small fraction of excitatory synapses must be specific in order to introduce selective amplification and competition within the network . Several previous models of recurrent cortical connectivity designed for columnar visual cortex have incorporated selective excitatory connectivity , either with connectivity relying on purely anatomical constraints ( e . g [79] ) or mimicking the spatially periodic , long-range lateral excitatory projections found in monkey , cat and other species ( e . g . [80–83] ) . Similarly to our models , these works emphasise that feature integration can occur within V1 through recurrent processing of visual stimuli . These earlier models examined how specific synaptic connectivity between spatially separated neurons across visual space can perform operations that link representations of similar visual features such as contour integration , and can underlie competition between dissimilar visual features [80 , 84] . The principal difference to our models is that we examined how local excitatory connections between neurons representing overlapping regions of visual space can underlie facilitatory binding of dissimilar visual features . Our models therefore examine the consequences of higher-order patterns in local recurrent connectivity on visual coding . In visual cortex of primates , carnivores and rodents , orientation tuning develops before postnatal eye opening and in the absence of visual experience [85 , 86] . Local recurrent connections develop after the onset of visual experience and maintain their plasticity into adulthood [85 , 87–91] . Statistical correlations in natural scenes might therefore lead to wiring of subnetworks under an activity-dependent mechanism such as spike-time dependent plasticity ( STDP ) [92–96] . Along these lines , examinations of the development of specific excitatory connections after eye opening found that similarities in feedforward input were progressively encoded in specific excitatory connections [22] . We expect that , as the specificity of lateral connections forms during development , the emergence of compound feature selectivity will gradually occur after the onset of sensory experience . This hypothesis is consistent with experience-dependent development of modulatory effects due to natural visual stimulation outside of the classical receptive field , as observed in mouse V1 [97] . A complete factorial combination of all possible features occurring in natural vision is clearly not possible . However , the most prominent statistical features of cortical activity patterns could plausibly be prioritised for embedding through recurrent excitatory connectivity . At the same time , competition induced by non-specific shared inhibition will encourage the separation of neurons into subnetworks . In our interpretation , single subnetworks would embed learned relationships between external stimulus features into functional ensembles in cortex , such that they could be recovered by the competitive mechanisms we have detailed . In prefrontal cortex , compound or mixed selectivity of single neurons to combinations of task-related responses has been found in several studies [98 , 99] . This is proposed to facilitate the efficient decoding of arbitrary decision-related variables . Binding feedforward cortical inputs into compound representations , as occurs in our feature-binding model , is therefore a useful computational process with general applicability . Our work suggests that specific local excitatory connectivity could be a general circuit mechanism for shaping information processing in cortical networks .
Experimental procedures followed institutional guidelines and were approved by the Cantonal Veterinary Office in Zürich or the UK Home Office . Procedures for urethane anesthesia , craniotomies , bulk loading of the calcium indicator , as well as for in vivo two-photon calcium imaging and in vitro recording of synaptic connection strength were the same as described previously [24 , 30 , 100 , 101] . Visual stimuli for receptive field characterisation , drifting gratings and plaids and masked natural movies were displayed on an LCD monitor ( 52 . 5 × 29 . 5 cm; BenQ ) placed 10–11 cm from the eye of the animal and covering approximately 135 × 107 visual degrees ( v . d . ) . The monitor was calibrated to have a linear intensity response curve . Contrast-oscillating grating and plaid stimuli were presented on an LCD monitor ( 15 . 2 × 9 . 1 cm; Xenarc ) placed 9 cm from the eye of the animal and covering 80 × 54 v . d . The same screen was used for stimulus presentation during intrinsic imaging to locate visual cortex and during two-photon imaging . The open-source StimServer toolbox was used to generate and present visual stimuli via the Psychtoolbox package [33 , 105] . Stimuli for receptive field characterisation comprised a 5 × 5 array of masked high contrast drifting gratings ( 15 v . d . wide; overlapping by 40%; 9 v . d . per cycle; 1 Hz drift rate; 0 . 5 Hz rotation rate ) presented for 2 s each in random order , separated by a blank screen of 2 s duration , with 50% luminance ( example calcium response shown in S1A Fig ) . Frames were averaged during the 2 s stimulus window to estimate the response of a neuron . Full-field high-contrast drifting gratings ( 33 . 33 v . d . per cycle; 1 Hz drift rate ) were presented drifting in one of 8 directions for 2 s each in random order , separated by a 6 s period of blank screen with 50% luminance ( example calcium response shown in S1B Fig ) . Frames were averaged during the 2 s stimulus window to estimate the response of a neuron . Full-field 50% contrast drifting sine-wave gratings ( 25 v . d . per cycle; 1 Hz drift rate ) were presented drifting in one of 16 directions for 1 s each in random order ( calcium responses shown in Fig 2 ) . Full-field drifting plaid stimuli were constructed additively from 50% contrast sine-wave grating components ( 25 v . d . per cycle; 1 Hz drift rate; 1 s duration; Fig 2 ) . Three frames were averaged following the peak response ( 384 ms window ) to estimate the response of a neuron . Full-field natural movies consisted of a 43 s continuous sequence with three segments ( example calcium response shown in S1C Fig ) . Full-field contrast-oscillating square-wave gratings and plaid stimuli were composed of bars of 8 v . d . width which oscillated at 2 Hz between black and white on a 50% grey background , and with a spatial frequency of 20 v . d . /cycle ( example calcium response shown in Fig 5A ) . On each subsequent oscillation cycle the bars locations shifted phase by 180° . Static gratings were used to avoid introducing a movement component into the stimulus . A base orientation for the gratings of either horizontal or vertical was chosen , and five orientations spanning ±40 deg . around the base orientation were used . Contrast-oscillating plaids were composed of every possible combination of the five oscillating grating stimuli , giving 5 grating and 10 plaid stimuli for each experiment . A single trial consisted of a blank period ( 50% luminance screen ) presented for 20 s , as well as presentations of each of the gratings and plaids for 5 s each , preceded by 5 s of a blank 50% luminance screen , all presented in random order . Frames from 0 . 25 s to 4 . 75 s during the stimulus period were averaged to estimate the response of a neuron . Analysis of two-photon calcium imaging data was conducted in Matlab using the open-source FocusStack toolbox [33] . During acquisition , individual two-photon imaging trials were visually inspected for Z-axis shifts of the focal plane . Affected trials were discarded , and the focal plane was manually shifted to align with previous trials before acquisition continued . Frames recorded from a single region were composed into stacks , and spatially registered with the first frame in the stack to correct lateral shifts caused by movement of the animal . Only pixels for which data was available for every frame in the stack were included for analysis . A background fluorescence region was selected in the imaged area , such as the interior of a blood vessel , and the spatial average of this region was subtracted from each frame in the stack . The baseline fluorescence distribution for each pixel was estimated by finding the mean and standard deviation of pixel values during the 10 s blank periods , separately for each trial . Regions of interest ( ROIs ) were selected either manually , or by performing low-pass filtering of the OGB ( green ) and sulforhodamine ( red ) channels , subtracting red from green and finding the local peaks of the resulting image . A general threshold for responsivity was computed to ensure that ROIs considered responsive were not simply due to neuropil activity . The responses of all pixels outside any ROI were collected ( defined as “neuropil” ) , and the Z-scores of the mean ΔF/F0 responses during single visual stimulus presentations were computed per pixel , against the baseline period . A threshold for single-trial responses to be deemed significant ( ztrial ) was set by finding the Z-score which would include only 1% of neuropil responses ( α = 1% ) . A similar threshold was set for comparison against the strongest response of an ROI , averaged over all trials ( zmax ) . These thresholds always exceeded 3 , implying that single-trial responses included for further analysis were at least 3 standard deviations higher than the neuropil response . Note that this approach does not attempt to subtract neuropil activity , but ensures that any ROI used for analysis responds to visual stimuli with calcium transients that can not be explained by neuropil contamination alone . The response of an ROI to a stimulus was found on a trial-by-trial basis by first computing the spatial average of the pixels in an ROI for each frame . The mean of the frames during the blank period preceding each trial was subtracted and used to normalise responses ( ΔF/F0 ) , and the mean ΔF/F0 of the frames during the analysed trial period was computed . The standard deviation for the baseline of a neuron was estimated over all ΔF/F0 frames from the long baseline period and the pre-trial blank periods . ROIs were included for further analysis if the ROI was visually responsive according to trial Z-scores ( maximum response > zmax ) and reliable ( trial response > ztrial for more than half of the trials ) . The response of a neuron to a stimulus was taken as the average of all single-trial ΔF/F0 responses . Receptive fields of neurons recorded under natural movie and drifting grating stimulation were characterised by presenting small , masked high-contrast drifting gratings from a 5 × 5 array , in random order ( see above; S1A Fig ) . A receptive field for each neuron was estimated by a Gaussian mixture model , composed of circularly symmetric Gaussian fields ( ρ = 7 . 5 v . d . ) placed at each stimulus location and weighted by the response of the neuron to the grating stimulus at that location . The centre of the receptive field was taken as the peak of the final Gaussian mixture . Neurons were included for further analysis if the centre of their receptive field lay within a 7 . 5 v . d . circle placed at the centre of the natural movie visual stimulus . Example single-trial and trial-averaged calcium responses to natural movie stimuli are shown in S1C Fig . The similarity in response between two neurons was measured independently for grating and plaid stimuli . The set of grating responses for each neuron were composed into vectors R1g and R2g , where each element of a vector was the trial-averaged response of a neuron to a single grating orientation . The similarity in grating responses between two neurons was then given by the Pearson’s correlation coefficient between R1g and R2g: ρg = corr ( R1g , R2g ) ( see S1B Fig , inset ) . The similarity in response to plaid stimuli was computed analogously over the sets of trial-averaged plaid responses R1p and R2p: ρp = corr ( R1p , R2p ) ( see Fig 5A , inset ) . Similarity was only measured between neurons recorded in the same imaging site . The similarity between neurons in their responses to movie stimuli ( ρm ) was measured by computing the signal correlation as follows . The calcium response traces for a pair of neurons were averaged over trials . The initial 1 s segment of the traces following the onset of a movie segment were excluded from analysis , to reduce the effect of transient signals in response to visual stimulus onset on analysed responses . The Pearson’s correlation coefficient was then calculated between the resulting pair of traces ( ρm; see S1C Fig , inset ) . Note that correlations introduced through neuropil contamination were not corrected for , with the result that the mean signal correlation is positive rather than zero . For this reason we used thresholds for “high” correlations based on percentiles of the correlation distribution , rather than an absolute correlation value . The similarity between neurons in their responses to flashed natural stimuli ( ρCa; S1F Fig ) was measured as the linear correlation between the vector of responses of a single neuron to a set of 1800 natural stimuli [24] . The Orientation Selectivity Index ( OSI ) of a neuron was estimated using the formula OSI = [max ( Rg ) −min ( Rg ) ]/sum ( Rg ) , where Rg is the set of responses of a single neuron to the set of grating stimuli . The OSI of a neuron ranges from 0 to 1 , where a value of 1 indicates that a neuron responds only to a single grating stimulus; a value of 0 indicates equal , nonselective responses to all grating stimuli . The Plaid Selectivity Index ( PSI ) of a neuron , describing how selective a neuron is over a set of plaid stimuli , was calculated using the formula PSI = 1−[−1 + ∑jRp , j/max ( Rp ) ]/[# ( Rp ) −1] where # ( Rp ) is the number of stimuli in Rp [30] . The PSI of a neuron ranges from 0 to 1 , where a value of 1 indicates a highly selective response , where a neuron responds to only a single plaid stimulus; a value of 0 indicates equal , nonselective responses to all plaid stimuli . A plaid Modulation Index ( MI ) , describing the degree of facilitation or suppression of a neuron in response to plaid stimuli , was calculated using the formula MI = [max ( Rp ) −max ( Rg ) ]/[max ( Rp ) +max ( Rg ) ] , where Rp is the set of responses of a single neuron to the set of plaid stimuli [30] . The MI of a neuron ranges from -1 to 1 . Values of MI < 0 indicate stronger responses to grating stimuli compared with plaid stimuli; values of MI > 0 indicate stronger responses to plaid stimuli . A value of MI = -1 indicates that a neuron responds only to grating stimuli; a value of MI = 1 indicates that a neuron responds only to plaid stimuli . The proportion of facilitating and suppressing neurons was compared between mouse V1 and model responses using two-tailed Fisher’s exact tests . The population of responsive neurons was divided into three groups: facilitating ( MI > 0 . 05 ) ; suppressing ( MI < -0 . 05 ) ; and non-modulated ( -0 . 05 < = MI < = 0 . 05 ) . These categories were arranged into three 2 × 3 contingency tables , with each table tested to compare facilitation and suppression between mouse V1 and one model . We used single-cell , single-trial responses to oscillating contrast grating stimuli to explore whether we could distinguish between correlated and decorrelated responses to plaid stimuli , given experimental variability and noise . For each cell in the experimentally-recorded data set , we used the set of grating responses Rg to generate plaid responses Rp for the same cell , under the assumption that the response to a plaid was linearly related to the sum of the responses to the two grating components . For each plaid , we randomly selected a single-trial response for each of the grating components of the plaid . The predicted single-trial plaid response was the sum of the two grating responses . We generated 100 bootstrap samples for each experimental population , with each sample consisting of the same number of trials and neurons as the experimental population . We then quantified the relationship between grating and plaid responses as described for the experimental data . We used a sample size commensurate with those used in the field , and sufficient for statistical analysis of our observations . No explicit sample size computation was performed . Two-sided , non-parametric statistical tests were used unless stated otherwise in the text . | The brain is a highly complex structure , with abundant connectivity between nearby neurons in the neocortex , the outermost and evolutionarily most recent part of the brain . Although the network architecture of the neocortex can appear disordered , connections between neurons seem to follow certain rules . These rules most likely determine how information flows through the neural circuits of the brain , but the relationship between particular connectivity rules and the function of the cortical network is not known . We built models of visual cortex in the mouse , assuming distinct rules for connectivity , and examined how the various rules changed the way the models responded to visual stimuli . We also recorded responses to visual stimuli of populations of neurons in anesthetized mice , and compared these responses with our model predictions . We found that connections in neocortex probably follow a connectivity rule that groups together neurons that differ in simple visual properties , to build more complex representations of visual stimuli . This finding is surprising because primary visual cortex is assumed to support mainly simple visual representations . We show that including specific rules for non-random connectivity in cortical models , and precisely measuring those rules in cortical tissue , is essential to understanding how information is processed by the brain . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"nervous",
"system",
"brain",
"social",
"sciences",
"electrophysiology",
"neuroscience",
"computational",
"neuroscience",
"vision",
"neuroimaging",
"neuronal",
"tuning",
"research",
"and",
"analysis",
"methods... | 2017 | Specific excitatory connectivity for feature integration in mouse primary visual cortex |
We study the structure and dynamics of spherical high density lipoprotein ( HDL ) particles through coarse-grained multi-microsecond molecular dynamics simulations . We simulate both a lipid droplet without the apolipoprotein A-I ( apoA-I ) and the full HDL particle including two apoA-I molecules surrounding the lipid compartment . The present models are the first ones among computational studies where the size and lipid composition of HDL are realistic , corresponding to human serum HDL . We focus on the role of lipids in HDL structure and dynamics . Particular attention is paid to the assembly of lipids and the influence of lipid-protein interactions on HDL properties . We find that the properties of lipids depend significantly on their location in the particle ( core , intermediate region , surface ) . Unlike the hydrophobic core , the intermediate and surface regions are characterized by prominent conformational lipid order . Yet , not only the conformations but also the dynamics of lipids are found to be distinctly different in the different regions of HDL , highlighting the importance of dynamics in considering the functionalization of HDL . The structure of the lipid droplet close to the HDL-water interface is altered by the presence of apoA-Is , with most prominent changes being observed for cholesterol and polar lipids . For cholesterol , slow trafficking between the surface layer and the regimes underneath is observed . The lipid-protein interactions are strongest for cholesterol , in particular its interaction with hydrophobic residues of apoA-I . Our results reveal that not only hydrophobicity but also conformational entropy of the molecules are the driving forces in the formation of HDL structure . The results provide the first detailed structural model for HDL and its dynamics with and without apoA-I , and indicate how the interplay and competition between entropy and detailed interactions may be used in nanoparticle and drug design through self-assembly .
Cardiovascular diseases are the primary cause of death in western countries [1] . One of the main causes is the lipid accumulation and plaque formation on arterial walls , called atherosclerosis . This eventually leads to the narrowing of arteries , plaque rupture , clotting , and potential death . Generally speaking , high levels of low density lipoprotein ( LDL ) in blood have been found to increase the risk of atherosclerosis [2] , [3] , whereas high levels of high density lipoprotein ( HDL ) have been shown to reduce the risk [4] , [5] . Despite more than a decade of extensive studies , LDL and HDL structures are not well understood . This is largely due to their small size which ranges from about 10 ( HDL ) to 25 nm ( LDL ) rendering experimental studies of the detailed lipoprotein structures extremely difficult . This challenge is further corroborated by the soft nature of lipoparticles whose structures are transient due to thermal forces driving molecular assembly processes in living matter . The challenge is to unravel the role and mechanisms of lipoproteins in the trafficking of cholesterol and in the cholesterol-based diseases . In this work , we focus on HDL . Let us briefly summarize the main insight one has about high density lipopproteins . HDL particles are comprised of a lipid droplet surrounded by proteins [6] , [7] . Apolipoprotein A-I ( apoA-I ) is the main protein associated with HDL , which is the main carrier of excess cholesterol from peripheral tissues to the liver , that is , for reverse cholesterol transport [5] , [8] . After synthesization , the ATP-binding cassette transporter A1 ( ABCA1 ) assembles lipid-poor apoA-I molecules and lipids into discoidal HDL particles [9] , after which the enzyme lecithin∶cholesterolacyl transferase ( LCAT ) esterifies cholesterol molecules , leading to the formation of spheroidal HDL [10] , [11] . The spheroidal HDL is the main form of HDL responsible for cholesterol transport to the liver . Though a number of experimental studies have been carried out to unravel the structure and dynamics of apoA-I molecules in lipid-free form [12]–[15] and in discoidal HDL complexes [16] , [17] , the structure of the spheroidal HDL has remained unclear . As for the structure of apoA-I , a large amount of data is in favor of the so-called double belt model ( see Ref . [18] and references therein ) , where the apoA-I proteins line along the lipid droplet . The composition of the droplet has been resolved [19] , [20] ( see Table 1 ) , indicating free cholesterol ( CHOL ) , cholesteryl esters ( CE ) , triglycerides ( TG ) , phospholipids , and lysolipids to be its main constituents , distributed such that there is a hydrophobic interior of triglycerides and cholesteryl esters and a surface covered by polar head groups of phospholipids [21] . This is essentially the so-called two-layer model for HDL [6] , [7] , [22] . Furthermore , parts of the apoA-I proteins have been proposed to interact with the acyl chains of the lipids [23]–[27] . Currently , the role of lipids for HDL functions are only vaguely understood . This is partly due to the transient time scales associated with and the nano-scale nature of HDL . Further issues include the poor understanding of lipid organization and interplay of lipids with apoA-I . Considering findings that lipids are an integral component of protein structures , e . g . , in membrane proteins that are in constant interplay with lipids [28] , it is obvious that clarifying the role of lipids for HDL properties is extremely important . A number of computational studies have recently been conducted to complement experiments . Previous computational studies of HDL particles have focused on discoidal particles consisting of phospholipids and two apoA-I molecules [29]–[32] . These studies have provided some insight into the mechanisms of assembly and the dependence of the particle shape on the lipid/protein molar fractions . In a different approach , bulk melts of cholesteryl esters [33] and triglycerides [34] , as well as combination of cholesteryl esters with POPCs [35] have recently been simulated . Catte et al . [36] reported the first computational approach towards understanding the structure of spheroidal HDL particles . Their study clarified the conformation of apoA-I in model spheroidal HDL particles using both all-atom ( AA ) and coarse-grained ( CG ) molecular dynamics ( MD ) simulations . This combination of AA and CG-MD simulations led to model spheroidal HDL particles with prolate ellipsoidal shapes having sizes consistent with experimental results and suggested that cholesteryl esters stabilize the conformations of apoA-I [37] . In a more recent work , Shih et al . also combined coarse-grained simulations with atomistic ones in a series of simulations where discoidal HDL was matured into spherical HDL upon incorporation of cholesteryl esters [38] . They found that maturation results from the formation of a dynamic hydrophobic core composed of cholesteryl esters , the core being surrounded by a layer of phospholipids and apoA-I proteins . Interestingly , Shih et al . also fine-grained the coarse-grained HDL particle to atomistic description and then used atomistic simulations to consider the structure of apolipoproteins around the lipid droplet , and in particular the importance of salt bridges in apoA-I . The main limitations of previous simulations of HDL particles are two-fold . First , the lipid composition modeled in recent simulations has been somewhat unrealistic: instead of a many-component lipid mixture , the lipid droplet has been modeled as a single-component POPC melt , or as a mixture of POPCs and cholesteryl esters [36] . The role of the many different lipid species in HDL has therefore remained unknown . Second , the time scales of HDL simulations have been too short compared to the characteristic time scales of lipid mixing and structural deformations associated with HDL . As even the time scale of lipid mixing is of the order of 1 s ( derived through the diffusion of lipids inside HDL ) , and the current state-of-the-art for atomistic simulations of HDL extends over 10–100 ns , it seems obvious that currently atomistic simulations are not the method of choice for dealing with HDL over large enough time scales . Our objective is to overcome the above limitations . We have performed MD simulations of spheroidal HDL particles using the full lipid composition of human plasma HDL [19] . Instead of atom-scale simulations , we employ the coarse-grained MARTINI model [39] , [40] that has performed exceptionally well in a number of studies dealing with lipids and proteins [36] , [39]–[42] . We consider both the protein-free lipid droplets and the full HDL particles containing also two apoA-I molecules around the droplet , see Figures 1 and 2 . Composition of the HDL system is described in Table 1 with abbreviations of all molecules included . By comparing the protein-free and the full HDL models , we can clarify the role of lipids and proteins in HDL . The principal objective is to fill the gap of detailed structural and dynamic information of lipids in spheroidal HDL particles . We also address questions related to the role of apoA-I proteins and their interactions with lipids in HDL structures . The currently incomplete understanding of the latter issue is largely due to the size heterogeneity of HDLs ( diameters range from 7 . 2 ( ) to 12 nm ( ) ) and the large flexibility of apoA-I . The latter renders the prediction of the positioning of different alpha helices of apoA-I on a spherical surface very difficult . The distribution of lipids inside HDL and their interplay with apoA-I are of profound interest . From a more general perspective , knowledge of the structure of spheroidal HDL is crucial for understanding the conformational changes when HDL makes the transition from discoidal to spheroidal shape , and the trafficking of CHOL and CE through the action of cholesteryl ester transfer protein [43] . Additionally , to design nanoparticles with desired surface and bulk properties , e . g . , for controlled transport and release of drugs and contrast agents , it is vital to understand the conformational changes as well as the underlying mechanisms in detail [44]–[51] .
The radial density distributions shown in Figure 3 reveal the internal structure of the simulated lipoparticles . The hydrophobic CE and TG molecules are located in the core of the particle and have minimal overlap with water . The lipids with a polar head group , POPC and PPC , are mostly located at the surface region facing water , whereas most of CHOL is located just below these two lipids . Note that a small but significant concentration of CHOL is also found in the core of the particle . Considering the size of HDL , the radii of gyration give an average of nm for the droplet and nm for the full HDL . Both particles are effectively spherical , as indicated by the moments of inertia ( data not shown ) . The apoA-I proteins are embedded onto the surface of the HDL particle , their density peaking just slightly below the headgroup region of POPC and PPC . The presence of the protein slightly disturbs the distribution of these lipids as revealed by the comparison of the lipid droplet with the full HDL particle . The distribution of hydrophobic lipids remains undisturbed . Most significant is the shifting of the distribution of CHOL , and partly PPC , towards water phase when the protein is present , while the distribution of POPC is shifted slightly towards membrane center , making room for CHOL and PPC . In the full HDL particle , water is found to distribute less to the particle compared to the droplet . Our results clearly highlight the displacement of CHOL even further towards the interface in the full HDL particle . The data below shows that CHOL interacts prefentially with some of the protein residues , strongly promoting the partitioning of CHOL to the vicinity of apoA-I . CHOL further prefers to reside next to the water region , facilitating ( hydrogen ) bonding via the polar OH group . It has been proposed [52] that CHOL molecules can mediate the relief of membrane stress arising from chain-chain mismatches , since their dimerization is not favored in membranes with high surface curvature . This view is supported by the observations of Huang and Mason [53] . Their work suggests that high surface curvature requires CHOL to be at the interfacial region . Interestingly , Lemmich et al . have further found that very small amounts of CHOL ( less than about 3 mol-% ) may soften the interface and hence promote its fluidity [54] . In HDL , the average concentration of CHOL is about 10 mol-% , but at the interface it is about 5–10 mol-% depending on distance from the water phase ( see Figure 3 ) . The minor but significant concentration of CHOL in the core of the particle calls for discussion . The usual assumption especially in studies of lipid membranes is that CHOL resides at the water-lipid interface due to its polar OH group . This is expected often to be the case , though there are also reported exceptions such as CHOL residing for short times in the middle of a polyunsaturated lipid bilayer [55] , [56] , and the suggestion of CHOL in the interior of LDL [57] . To start with , one gets an impression that the density plot adheres to the two-layer model [6] , [7] , [22] wherein one assumes almost full separation of hydrophilic and hydrophobic molecules into two separate regions . While the distribution of TG fits into this picture , the distribution of CE and CHOL does not . A rather significant amount of CHOL is also in the core of the particle as was discussed above . Detailed consideration further reveals that there is a significant overlap of CE with CHOL , POPC , and PPC: The radial density distributions shown in Figure 3 do not provide a sufficiently unique description of only two different structural regions inside HDL . Furthermore , by looking at the order parameters of CHOL and CE presented in Figure 4 it becomes evident that there are not only two regions but also the intermediate one between the hydrophobic and hydrophilic ones . The innermost core of the particle ( nm ) is clear , as there the ring structures of both CHOL and CE are oriented in a completely random fashion . The situation changes as one approaches the lipid-water interface through the intermediate region ( 3 nm nm ) , which is characterized by significant ordering of the ring structures , in a manner where the principal axis of CE's and CHOL's ring moiety lies along the radial direction of HDL . This intermediate region overlaps with the distribution of the acyl chains of POPC and PPC , revealing that the sterol rings are also aligned with the acyl chains . Finally , at the HDL-water interface , one finds the region composed of hydrophilic headgroups of POPC and PPC that constitute the surface part of the lipid droplet interacting mostly with water . The data clearly shows that instead of the two-layer model , the distribution of lipids in HDL is more complex . The structural nano-scale organization of CHOL and CE plays an important role in constituting the intermediate layer . However , there is no apparent reason to conclude that the lipid droplet in HDL would be described by a “three-layer” model either , since the intermediate region is narrow and represents a crossover from the hydrophobic to the hydrophilic environment rather than a clearly defined layer of its own . Our results for lipid dynamics are in favor of this view and will be discussed below in the context of diffusion . Meanwhile , while quantitative results have been missing , a three-layer model has earlier been proposed for LDL particles [3] . There the situation is different , though , since the diameter of LDL is roughly three times larger compared to HDL and the intermediate region can possibly be broader and more characteristic compared to HDL . There are significant differences when the order parameters of CHOL and CE are compared ( Figure 4 ) . First , the height of the main peak is considerably lower for CE than for CHOL , indicating that the ring of CE has a lower tendency to orient itself along the acyl chains than CHOL . Second , unlike for CHOL , on the surface of the particle ( nm ) the order parameter of CE obtains negative values . These indicate the ring of CE to lie along the surface , perpendicular to the radial direction . This obviously stems from entropic reasons , since while CE strives in part to organize its structure like CHOL , also directing its weakly polar ester bond region to the surface like CHOL does for the OH group , CE also has a long oleate chain . Previous atomistic simulations of CE in bulk conditions as well as in a combined system with POPCs have shown that the oleate chain of cholesteryl oleate has essentially three different conformations with respect to the ring of CE [33] , [36]: one where the angle of the oleate chain ( describing it as a semi-stiff rod ) with respect to the principal axis of the ring is about 35 degrees , and two other modes with an angle of 90 or 150 degrees . Compared to CHOL with only one mode , CE inevitably aims to minimize free energy by promoting entropic degrees of freedom . Another interesting observation is that apoA-I suppresses the main peak of both CHOL and CE molecules in Figure 4 . The effect is an indication that the protein disturbs the ordering within the intermediate region ( between the core and the surface ) , also facilitating the displacement of CHOL towards the water phase . This conclusion is supported by the broadening of the angle distributions of POPC in the presence of the protein ( see Supporting Information ( SI ) ) . An analysis of the internal conformations of CE molecules in Figure 5 provides a more detailed view of the situation . In the core of the particle , the most probable conformation of CE is the coil-like conformation ( maximizing entropy ) , where the angle between the CE ring and the oleoyl chain is about 120 degrees . This is largely consistent with recent atom-scale simulations of CE in bulk conditions [33] . The behavior changes on the surface of the particle . The two peaks of the distribution on the surface correspond to two distinctly different conformations: one where the ester group of CE ( corresponding to the OH-group of CHOL ) points towards water and the oleate chain is extended towards the solvent , and another where the ester region is pointing towards the core of the particle , while the ring and the oleoyl chain form a small angle with each other . Also for TG , we find a change of conformation when it is shifted from the core of the particle onto the surface . In the core , the three chains of TG place themselves to a similar conformation as in a bulk melt of TG [34] . When brought to the surface , the ester bond regions seek contact with water , which brings the three chains of TG closer to each other into a more tightly packed conformation ( see Figure S2 ) . Additional data for molecular conformations are presented in Figure S1 , Figure S3 , Figure S4 , and Figure S5 . The large-scale dynamics within HDL and the lipid droplet are considered in terms of diffusion , characterized by the diffusion coefficient . The diffusion coefficients were determined by considering lipid displacement distribution functions over a fixed period of time ( see SI ) . We found that the jump length distributions for lipids in the core of the particle ( TG and CE ) fitted well with the three-dimensional diffusion model , yielding . Meanwhile , the lipids on the surface ( POPC , PPC , CHOL ) fitted much better with the two-dimensional description for diffusion , yielding . For details , see SI . Table 2 shows the long-time diffusion coefficients of the lipid components within the lipid droplet and the full HDL particle . Figure 6 depicts how the diffusion rate varies significantly inside the lipid droplet and/or full HDL . The diffusion is the slowest in the middle of the particle , it speeds up as the molecules get closer to the interface , and the fastest diffusion is found at the interface . The influence of apoA-I on diffusion of lipids is modest . It turns out that the lipid diffusion coefficients in the protein-free lipid droplet and the full HDL particle are almost similar . The apoA-I proteins may slow down the diffusion of lipids slightly especially close to the interfacial regions . The effect is , however , weak ( see Table 2 ) . The diffusion coefficients of POPC , PPC and CHOL in the surface region of the particles are about and in good agreement with experimental estimates of for two-dimensional lipid bilayers in fluid phase [58] . On the other hand , the diffusion coefficients for CE and TG are smaller by a factor of 10 , about . To our knowledge , diffusion coefficients of lipids in HDL have not been experimentally determined . However , for LDL and LDL-like lipid droplets , Vauhkonen et al . used pyrene-linked PC lipids as probes to find that at the surface of lipoparticles [59] , in good agreement with our findings . Massey and Pownall have further used single-chain cationic amphiphiles for considering lipid mobility at the surface regions of LDL and HDL , and while quantitative estimates for are missing , they concluded that the diffusion at the surface is about 2–3 times slower compared to cholesterol-free POPC vesicles [60] . Recent MARTINI-model simulations for single-component PC bilayers have yielded [61] , which is about a factor of 2 larger than diffusion at the surface of HDL . While the comparison of our simulation data and experiments is suggestive rather than conclusive , the qualitative agreement is striking . Our main result regarding diffusion is that diffusion at the surface region of HDL is largely similar to diffusion of lipids in cholesterol-containing lipid bilayers in the fluid-like phase , the cholesterol concentration being roughly 10 mol% . Figure 6 also shows convincingly that the effect of apoA-I on diffusion of lipids is not significant . Additionally , Figure 6 provides compelling evidence that the dynamics of lipids in terms of their diffusion properties is not consistent with the two-layer model . Instead of two clearly different dynamic regions , we find the diffusion coefficients to increase monotonously: diffusion rates are clearly different in the core ( nm ) , in the ordered intermediate region ( nm ) , and at the surface ( nm ) . Given the different proposed models for lipid distribution in HDL , the striking difference of the present findings compared to earlier studies is the role of dynamics: not only the structural and ordering properties of molecules in HDL differ across HDL , but also the dynamics in terms of molecular transport coefficients varies significantly in the different compartments . The biological relevance of this feature lies in the time scales of molecular trafficking inside HDL: while molecular transport between the surface and the intermediate region is relatively fast , the transport between the surface and the core of HDL is slower by a factor of 10 . The above results show that the apoA-I proteins do not induce large changes to the lipids' properties inside the droplet . Yet , while the protein collapses onto the surface of the droplet , it does disturb the packing , ordering and , although only slightly , also the dynamics of the lipids at the surface region . What remains to be explored is the nature of the lipid-protein interactions . In this work our primary interest is the lipid component of HDL , thus we have used the standard CG MARTINI model which does not enforce the full secondary structures in apoA-I . This optimizes computational efficiency and allows us to focus on generic issues such as the partitioning of lipids around apoA-I , and the influence of apoA-I on the lipid droplet . Meanwhile , we cannot address questions related to detailed atomistic phenomena at the lipid-apoA-I interface . Data for the surface accessible surface areas ( SASAs ) of apoA-I hydrophilic and hydrophobic residues ( data not shown ) provide evidence for the low contribution of protein hydrophobic residues ( 11% ) to the total SASA of the protein , the main contribution coming from protein hydrophilic residues ( 89% ) . The average value of SASA of protein hydrophobic residues ( ) is in good agreement with that reported by Shih et al . [32] in a recent study on the assembly of lipids and proteins into lipoprotein particles . The RMSFs of protein carbons are shown in Figure 7 and reveal the mobility of different protein domains . The -helical structure of the protein exhibits very little mobility for both chains . This rigidity of the protein is also in agreement with the observed slight disturbances produced on the lipid packing . The number of annular lipids , as defined in the Method section , is given in Figure 8 for each lipid component . It is interesting to note that about 80% of CHOL molecules are annular ( on average 40 out of 49 ) while only 10% of CE molecules ( about 15 out of 122 ) are in close contact with the protein . An average of about 98 POPC molecules out of a total of 260 are annular . Overall , the results indicate that there is a preferential interaction between CHOL molecules and protein residues . This result is striking if one considers that the number of POPC molecules is larger than that of CHOL molecules . It is known that the number of apolipoproteins in HDL depends on particle size . We characterized its role for lipid distribution through additional simulations with three apoA-Is . First , we performed a 20 microsecond simulation of the same lipid droplet with 3 apoA-I molecules placed 2 nm apart from each other . The protein molecules were found to insert themselves in the lipid droplet in the same way as was observed above , with hydrophobic moieties pointing towards the droplet . The only interesting difference was that in the structure with three apoA-Is , the C-terminus and the helix 9 of one protein molecule were not inserted in the lipid droplet . This situation is likely due to the crowded arrangement of apolipoproteins , or due to the limited time scale of the simulation . The addition of a third apoA-I molecule does also affect the interaction of cholesterol with apoA-I: Almost 100% of the cholesterol molecules ( out of 49 ) are in contact with the three proteins . That is , the addition of the third apoA-I molecule enhances the average number of annular cholesterol molecules from about 80% ( observed with 2 apoA-I molecules ) to about 96% of the total unesterified cholesterol in the particle . The lipid-protein interactions of different moieties of each lipid component showed that POPC , PPC and TG molecules interact with apoA-I residues preferentially through the acyl chains ( POPC and PPC glycerol backbone has also a high number of contacts with apoA-I ) , while CHOL and CE molecules interact with the protein mainly through the sterol ring ( see Table S1 ) . To better understand the nature of the interaction between CHOL and apoA-I we also measured the number of lipid-protein contacts per residue ( hydrophobic and hydrophilic ) , shown in Figure 9 . It is clear that there is a preferential interaction of CHOL molecules with apoA-I hydrophobic residues , in particular tryptophane ( Trp ) and phenylalanine ( Phe ) having aromatic side chains , but also valine ( Val ) and leucine ( Leu ) . Highly preferred interaction with Trp and Phe is understandable through findings of aromatic ring pairing in e . g . known protein structures [62] . We also observe a relevant number of contacts with apoA-I hydrophilic residues with aromatic side chains such as tyrosine ( Tyr ) and histidine ( His ) . This is not surprising , as Tyr has a hydrophobicity comparable to Phe as has been shown experimentally by Wimley and White [63] through the determination of a hydrophobicity scale for proteins at membrane interfaces . There are less contacts of CHOL molecules with the other apoA-I hydrophilic residues , namely serine ( Ser ) , threonine ( Thr ) and asparagine ( Asp ) being the most attractive ones . These results are in good agreement with the observed large number of contacts of the sterol ring of CHOL molecules with protein residues . The sterol ring of CHOL molecules can intercalate or interact with the aromatic side chains of protein residues as observed for CE in a recent study by Catte et al . [36] . This interaction between CHOL molecules and apoA-I was also observed experimentally by Dergunov et al . [64] . The authors estimated the degree of exclusion of CHOL molecules from the boundary lipid region in reconstituted discoidal HDL particles containing different apolipoproteins and observed an increase in the order A-I<E<A-II . The partial exclusion of CHOL molecules operated by apoA-I and the corresponding CHOL distribution among surface and bulk lipids are in good agreement with our findings as depicted through a series of snapshots in Figure 10 ( see also SI ) . The binding between CHOL molecules and apoA-I residues is quite weak , which permits exchange among apoA-I -bound and free CHOL molecules on the time scale of the simulation . We characterized this trafficking process by computing the distributions of lifetimes between CHOL-protein and CE-protein contacts . The average lifetime was found to be 146 ns for CHOL and 15 ns for CE . While the errors are of the same order as the lifetime due to a limited number of samples , and the fact that the distribution for CHOL is broad as there are cases where the CHOL-protein contact is maintained throughout the simulation , the results highlight the stability of CHOL-protein binding with respect to that of CE . The relatively large lifetime of the CHOL-protein binding also highlights that once CHOL has migrated to the vicinity of apoA-I , it remains there for a long period of time . For comparison , the average non-contact lifetime for CHOL-protein pairs , describing the characteristic time for CHOL to not be in contact with any parts of apoA-I was found to be about 175 ns . That is , CHOL molecules reside close to the water-HDL interface and on average spend half of their time in contact with apoA-I . The above results are in good agreement with NMR experiments performed on human HDL , which indicate that CHOL molecules are present in two distinct environments [65] . More specifically , Lund-Katz et al . found that the cholesterol molecules dissolved in the core of HDL are relatively disordered and mobile , while the cholesterol molecules located among phospholipid molecules in the surface of the particle undergo relatively restricted , anisotropic motions . This view is in line with our simulation results discussed earlier in this article . Lund-Katz et al . thus proposed that cholesterol molecules are in two different microenvironments , undergoing fast exchange between these two regions , equilibrating between the surface and the core of HDL in the time scale of milliseconds or more . While the time scales proposed by Lund-Katz et al . are beyond those that are accessible via simulations , we have found that there is local exchange taking place at times up to microseconds . However , the time scales we have found via simulations should be regarded as the lower limit , since the diffusion coefficients we have found for the core of HDL imply that the exchange of cholesterols between the core and surface regions has to be larger than 1 s .
In this study , we elucidated the structure and dynamics of spheroidal high density lipoparticles with a realistic lipid composition corresponding to human serum HDL . We found that the traditional two-region model for HDL is not accurate enough . Instead , we found the distribution of the different lipid types in HDL to be more complex . The innermost core of HDL is mainly occupied by TG and CE , which as hydrophobic lipids constitute a randomly oriented melt . However , in contrast to the common view , the inner core was also found to contain a rather significant fraction of free cholesterol partitioned into the disordered melt . The outermost surface region constitutes the interface with water , which is mostly occupied by the polar headgroups of POPC and PPC . Between these two is the intermediate region occupied by CHOL , partly also CE , and the acyl chains of POPC and PPC . However , there is no apparent reason to consider the intermediate region as a “third layer” , since it is narrow , unclear to define spatially , and represents a crossover from the hydrophobic to the hydrophilic environment rather than a true layer of its own . Yet it has properties that are distinct from those in the core and at the surface . This is most obvious in two aspects: ordering of steroid moieties and molecular diffusion . Unlike in the core , in the intermediate region the bulky rings of CHOL and CE are strongly ordered along with the acyl chains . This ordering extends also to the surface region of HDL , highlighting the difficulty to define the intermediate region as a true layer of its own . This view is also supported by the diffusion data , which illustrates that the diffusion of molecules takes place at a clearly different pace in the different regions . In the core and in the intermediate region of the particle , diffusion was found to be three-dimensional , while the diffusion of lipids at the HDL-water interface turned out to be two-dimensional in nature . Quantitatively , diffusion in the core of the particle was observed to be slow , as in a polymer melt , and to speed up monotonously as one crosses the intermediate region and ends up in the interfacial region . The perspective arising from our results is novel , providing the first molecular scale view to the nano-scale organization of lipids in HDL . The present results indicate that the spatial distribution of lipids within HDL provides only a narrow perspective to the complexity of lipid organization . To understand this issue , one has to pay considerable attention not only to density distributions but also conformational and orientational degrees of freedom of the lipids , and their dynamics within HDL . Events where CE and TG penetrate to the surface were found to be rare . In the few observed cases when it occurred , their conformation was significantly changed . In the core of the particle , both CE and TG were observed to be capable of obtaining more coil-like conformations . This indicates that the formation of the HDL core is not only driven by the hydrophobic effect , but that conformational entropy has a significant effect . When comparing the simulation of the full HDL ( with apoA-I ) to the lipid droplet ( without apoA-I ) , we found that the overall structure of the lipid droplet was not significantly changed by the presence of the protein . Rather , we found a disturbance in the behavior of the surface lipids . In particular , the order of CHOL and CE molecules decreased and the conformations of the acyl chains of PC lipids got broader . Diffusion of the surface lipids was slightly perturbed by the protein , but the effect was minor . The low contribution of the SASA of apoA-I hydrophobic residues to the total SASA of the protein and the large number of contacts of hydrophobic moieties of each lipid component with apoA-I evidence that the hydrophobic forces drive the insertion of the protein and contribute to the stability of the full HDL . Interestingly , we found that a large number of CHOL molecules interact with apoA-I , mainly through their sterol ring and especially with hydrophobic residues having an aromatic side chain . We also observed fast exchange among protein-free and protein-bound CHOL molecules . This result is in good agreement with experimental findings for human HDL particles [65] . It is tempting to discuss the physiological relevance of the above-discussed molecular level findings , especially the preferable interaction of the sterol ring moiety with the aromatic amino acids , and the observation that CHOL molecules spend about half of their time in contact with apoA-I , trafficking relatively rapidly back and forth in the vicinity of apoA-I . Such interaction is prone to have impact on the availability of sterols and lipids for related transfer proteins and enzymes , such as the cholesteryl ester transfer protein ( CETP ) and cholesteryl esterases . This interaction may be even more important in the process of cholesterol efflux , which is the critical part of reverse cholesterol transport , where the accumulated cholesterol is removed from macrophages . For future purposes for characterizing the properties of HDL , as well as HDL under enzymatic reactions , our results bring about a useful view to consider . We have found that the interfacial region of HDL close to the water phase is rather well defined in terms of its molecular composition: it can be described as a model layer composed of PCs , lyso-PCs , CHOLs , and the apolipoproteins A-I . The diffusion results discussed in this study indicate that the lateral diffusion along the interfacial layer is largely consistent with diffusion taking place in model membranes , whose molecular composition is of the same type . These features suggest that both the physical and chemical properties of HDL could be explored with reasonable accuracy through studies of ( planar ) model membranes , which are considerably easier to characterize compared to nano-sized HDLs . Clearly , the insight gained in this manner would be limited , since a number of inherent features would be missing , such as the curved nature of the HDL-water interface and its effect on apoA-I . Nonetheless , there is reason for optimism , encouraging experiments and simulations to use model membranes for better understanding of lipoprotein properties , including both HDL and LDL . The view presented in this article for HDL structure and dynamics paves way to extend the scope of computational studies for HDL , and to gain a much deeper understanding of HDL properties in a number of conditions related to health . For instance , there is reason to assume that the molecular composition in HDL depends to some extent on factors such as diet and lifestyle . In altered HDL the lipid composition can be abnormal due to e . g . dyslipidemia [66] . Characterization of molecular composition of HDL of subjects with varying degrees of health would allow coarse-grained simulation studies of HDL in these subject groups , using the present results as a reference . Preliminary studies in this spirit have very recently been reported and discussed by Yetukuri et al . [67] , who found that an elevated triglyceride concentration in low-HDL subjects also affected its distribution in HDL , increasing the concentration of triglycerides markedly at the lipid-water interface next to apolipoproteins . Such results based on large-scale coarse-grained models can further be fine-grained to atomistic description to study the atom-scale features that are relevant e . g . in lipid-protein interactions , and the implications on HDL stability due to reactions of enzymes such as phospholipases . Work in this direction has already been initiated by Shih et al . , who recently fine-grained coarse-grained models for matured HDL particles comprised of apoA-Is , phopholipids , and cholesteryl esters [38] . Similar work is in progress for the present HDL models . Altogether , considering the complexity of HDL , our study highlights the importance of lipid-apoA-I interactions , and in particular the specificity of apoA-I for free cholesterol and its esters . The molecular-scale insight of HDL structure and dynamics found and confirmed in our study largely stems from the ordering and dynamical phenomena taking place close to the HDL-water interface , being in part driven by the interactions between cholesterol and apoA-I . Evidently , they have a prominent role to play in a number of transport processes dealing with cholesterol .
Construction of the models was implemented in two stages . First , we constructed lipid droplets ( without the apoA-I proteins ) using coarse-grained descriptions of lipids and water . Second , the studies of pure lipid droplets were complemented by models where the droplet was surrounded by two apoA-I proteins . Below , we describe the main stages of the model construction . The initial structure for the lipid droplet was obtained by placing the set of lipid molecules ( see Table 1 ) randomly into a three-dimensional simulation box without water or any other solvent . As CE , we used cholesterol oleate , while TG was chosen as trioleate . The system was then simulated under NpT conditions in order to reach proper density . The resulting molecular melt was next hydrated with 100 , 000 water particles and the energy was minimized , after which the system was equilibrated for 8 s . After equilibration , the system was simulated over a period of 4 s that was used for analysis . All time scales shown here represent the realistic effective time ( simulation time multiplied by the scaling factor of four ) [39] . For POPC , PPC , CHOL and water , we used standard components of the coarse-grained MARTINI force field [39] . The parameters for CE and TG are those corresponding to cholesteryl-oleate and trioleate respectively , and they come from a combination of standard MARTINI-components and careful adjustments of the key particle types and angle potentials ( see SI ) . The adjustments were justified by comparing structural properties of the molecules in bulk with extensive atomic-scale simulations [33] , [34] , see SI for details . Sphingomyelin ( SM ) in ref . [19] has been included in the POPC contribution , as SM's properties in the MARTINI description do not differ considerably from those of POPC . Next , all-atom apoA-I molecules were generated using as a reference the molecular belt model of apoA-I for discoidal HDL [68]–[72] . The hydrophobic faces of the amphipathic helices were oriented toward the interior of the alpha helical ring but for the N-terminal part of apoA-I; the first 32 residues of the N-terminus were rotated , as in the lipid-mimetic solution NMR structure of apoA-I [22] , [73] , in order to have their hydrophobic face oriented towards the lipid droplet surface . These all-atom models of apoA-I were coarse grained using a pre-released version of the MARTINI force field for proteins [40] for the assignment of beads to every amino acid residue ( see also ref . [36] ) . To obtain the initial structure of the full HDL , two coarse grained apoA-I molecules were added to the equilibrated lipid droplet ( discussed above ) in a double-belt conformation at a distance of 4 nm from each other . After energy minimization , the HDL particle was subjected to very short equilibration runs using different time steps to get a stable system for a simulation with a time step of 25 fs . Finally , the particle was simulated for a total of 19 s , of which the last 4 s was used for analysis . To confirm the validity of the results , the simulations for HDL discussed in this article were complemented by several additional simulations that were started from different initial configurations . Each simulation covered a multi-microsecond time scale , and the results were found to be consistent with those discussed in this article . The molecular dynamics simulations were performed with the GROMACS 3 . 3 . 1 package [74] . Time steps of 20 fs and 25 fs were used for integrating the equations of motion of the lipid droplet and of the full HDL , respectively . For production runs , the Nosé-Hoover thermostat [75] , [76] and the Parrinello-Rahman barostat [77] were used to ensure proper NpT conditions ( K , atm ) . Water and the lipids were coupled to separate thermostats , and the whole system was coupled to the barostat isotropically . Time constant of ps was used for all couplings . For non-bonded interactions , we used the standard distance of 1 . 2 nm [36] . The Lennard-Jones interaction was shifted smoothly to zero after 0 . 9 nm . The equilibration of the simulated lipoparticles was monitored through the numbers of intermolecular contacts between different lipid types and the radial density distributions as a function of time . The intermolecular contacts between different molecular groups were calculated using a 0 . 8 nm cutoff for all beads . The radial density distributions describe the number densities of the coarse-grained beads . The orientation order of CHOL and CE ring structures was measured by the order parameter , where is the angle between the molecular axis and the effective normal of the lipoparticle at the location of the molecule in question . Being more specific , the molecular axis in this definition for CHOL is drawn from the beginning of CHOL ( carbon in the ring of CHOL attached to the short chain ) to the carbon connected to the hydroxyl group . The effective normal is the vector from the center of mass ( COM ) of the lipoparticle to the center of the molecular axis . For CE , an additional measure is the angle between the molecular ( ring , see above ) axis and the vector from the beginning to the end of the oleoyl chain . Diffusion was analyzed by measuring the jump-length distributions of the COM positions of the lipids over a time scale . Two types of Gaussian functions were fitted to the distributions , the two-dimensional ( 2d ) :and the three-dimensional ( 3d ) : The diffusion coefficient from the best fitting function ( or ) is reported , which in practice means that lipids at the water-lipid interface were found to undergo 2d diffusion , while those under the interface diffused in a 3d manner . Also , different time scales were tested and the measured was observed to level off at long times , an indication of diffusive behavior in the hydrodynamic ( long-time ) limit . “Long” times here refer to times of the order of 100 ns , where is found to level off to a well defined constant value . Examples of data for and are shown in SI , including also a more detailed description of how to choose the diffusion time scale in the intermediate region under the lipid-water surface ( see Text S1 , and Figure S6 ) . In many-component systems such as the present one , the diffusion of different molecular components may take place at different rates , and it is not obvious that CG models account for this aspect correctly . For the MARTINI model used here , we have previously confirmed that this is not an issue . For instance , Niemela et al . [78] recently used atomistic and coarse-grained models to show that the protein diffusion coefficient was about 10 times smaller compared to that of lipids , and the diffusion mechanisms of lipids and proteins was similar in both models . Ramadurai et al . [79] studied the influence of membrane thickness ( hydrophobic mismatch ) with several peptides using both FCS measurements and coarse-grained simulations and found essentially quantitative agreement for the peptide diffusion coefficients after the MARTINI results had been scaled by a factor of 4 . The simulations were also in agreement with experiments for the trend predicted with increasing hydrophobic mismatch . Further , Apajalahti et al . [61] considered the lateral diffusion of lipids in many-component protein-free membranes and found the diffusion of lipids in raft-like membrane domains ( in the liquid-ordered phase ) to be about 10 times slower compared to diffusion in domains that were in the liquid-disordered phase . Therefore , diffusion has been studied in several multi-component lipid systems , and the MARTINI models have been found to be consistent with experiments . The solvent accessible surface area ( SASA ) of hydrophobic , hydrophilic and all-protein residues were measured using a radius of the solvent probe of 0 . 56 nm ( the all-atom 0 . 14 nm radius of the solvent probe is converted to the coarse-grained one because one water bead corresponds to four water molecules ) inside the GROMACS program gsas . The SASA values were averaged over the entire trajectory used for analysis . The root mean square fluctuations ( RMSFs ) of protein alpha carbons were measured for both apoA-I chains to monitor protein flexibility . Lipid-protein interactions were monitored through the number of intermolecular contacts and their lifetimes for every lipid component with the protein . Annular lipid molecules , defined as those with any bead within 8 Å of any protein bead , were monitored over the analyzed trajectory . The average percentage of the number of contacts of each lipid component with the protein residues were estimated separately for each of the following moieties of every lipid component: POPC ( polar head group , glycerol backbone , oleoyl and palmitoyl chains ) , PPC ( polar head group , glycerol backbone and palmitoyl chain ) , CHOL ( short acyl chain , sterol ring ) , CE ( short acyl chain , sterol ring , and oleate chain ) and TG ( glycerol backbone , sn–1 , sn–2 , and sn–3 chains ) . Cholesterol-protein interactions were also tested by measuring the average number of contacts per protein residue of the cholesterol molecule with hydrophilic and hydrophobic protein residues . Additionally , for evaluation of CHOL-protein and CE-protein lifetimes , we accounted for cases where the distance between the molecules fluctuated around 8 Å: for an annular lipid , if its distance from apoA-I exceeded 8 Å temporarily for less than 10 frames ( 0 . 1 ns ) , the coupling was considered unbroken . Here our primary interest is the lipid part of HDL , for which reason we have used the standard CG MARTINI model which does not enforce the full secondary structures in apoA-I . This computational efficient approach allows us to focus on generic issues such as the partitioning of lipids around apoA-I , as well as the influence of apoA-I on the disttributions of lipids in a droplet . By fine graining our equilibrated structures back to atomistic level , one could employ atom-scale simulations to elucidate the more detailed aspects of the system . | Cardiovascular diseases are the primary cause of death in western countries . One of the main causes is lipid accumulation and plaque formation on arterial walls , called atherosclerosis . The risk of being exposed to this condition is reduced by high levels of high density lipoprotein ( HDL ) . The functionality of HDL has remained elusive , and even its structure is not well understood . Through extensive coarse-grained simulations , we have clarified the structure of the lipid droplet in HDL and elucidated its interactions with the apolipoprotein A-I ( apoA-I ) that surrounds the droplet . We have found that the structural and dynamic properties of lipids depend significantly on their location in the particle ( core , intermediate region , surface ) . As for apoA-I , we have observed it alter the overall structure of the lipid droplet close to the HDL-water interface , with prominent changes taking place for cholesterol and other polar lipids . The nature of lipid-protein interactions is most favorable for cholesterol . Our results reveal that not only hydrophobicity but also conformational entropy are the driving forces in the formation of HDL structure , suggesting how the interplay and competition between entropy and detailed interactions may be used in nanoparticle and drug design through self-assembly . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"computational",
"biology/molecular",
"dynamics",
"biophysics/macromolecular",
"assemblies",
"and",
"machines",
"biochemistry/macromolecular",
"assemblies",
"and",
"machines",
"biophysics/theory",
"and",
"simulation"
] | 2010 | Role of Lipids in Spheroidal High Density Lipoproteins |
Normal variation in gene expression due to regulatory polymorphisms is often masked by biological and experimental noise . In addition , some regulatory polymorphisms may become apparent only in specific tissues . We derived human induced pluripotent stem ( iPS ) cells from adult skin primary fibroblasts and attempted to detect tissue-specific cis-regulatory variants using in vitro cell differentiation . We used padlock probes and high-throughput sequencing for digital RNA allelotyping and measured allele-specific gene expression in primary fibroblasts , lymphoblastoid cells , iPS cells , and their differentiated derivatives . We show that allele-specific expression is both cell type and genotype-dependent , but the majority of detectable allele-specific expression loci remains consistent despite large changes in the cell type or the experimental condition following iPS reprogramming , except on the X-chromosome . We show that our approach to mapping cis-regulatory variants reduces in vitro experimental noise and reveals additional tissue-specific variants using skin-derived human iPS cells .
The recent advances in whole genome association studies ( GWAS ) have uncovered multiple genetic loci linked to common human diseases and traits . In addition to the more interpretable coding sequence changes , a large number of identified loci are in the non-coding region , suggesting that inheritable regulatory polymorphisms may play an important role [1]–[3] . Expression quantitative trait loci ( eQTL ) studies can reveal both cis- and trans-regulatory variants that can be mapped to a specific genetic region [4] , [5] . However , it requires a large sample size to reach the statistical power necessary to observe subtle changes in gene expression due to noise , ‘batch effects’ and other confounding factors [1] , [6] . Current mapped eQTL loci account for only a small fraction of the overall genetic risk for a given trait , suggesting that the weak effects from multiple genetic loci may play an important role . Although eQTL loci in different tissues can overlap [7]–[11] , the range of cell types available for study still poses a problem since many regulatory pathways are tissue-specific [1] , [12] . Given the potential of eQTL for elucidating genetic causes of complex traits and diseases , an ambitious effort has been launched to collect various tissue types from a large number of individuals ( i . e . Genotype-Tissue Expression project ) . However , the existing approaches to tissue sampling , including the use of surgical and tumor specimens , are complicated by social , medical and legal issues in addition to artifacts associated with tissue collection and processing [1] . In addition , it is difficult to follow up with a functional assay in the same individual and evaluate the biological effect of regulatory variants in the absence of a viable experimental system ( i . e . cell lines ) . Induced pluripotent stem cells [13]–[16] can be derived from skin , hair or blood [17] , [18] , using transduction of reprogramming factors ( i . e . OCT4 , SOX2 , KLF4 and MYC ) . They can be used to derive a number of tissues and cell types in vitro without resorting to invasive biopsy , and differentiation of iPS cells can theoretically allow for tissue-specific eQTL studies . However , the difficulty in observing pure and/or consistent in vitro differentiation can result in significant experimental variability and mask subtle regulatory variants given the practical limits on the sample size . An alternative approach may be to compare the expression level between two heterozygotic parental genes using ‘reporter’ SNPs ( expression SNPs ) in the exon [19]–[26] . Allele-specific gene expression ( ASE ) results from cis-regulatory differences in transcription ( i . e . upstream activating sequences , DNA methylation , core promoters ) or processing ( i . e . alternative splicing , miRNA ) [27] , [28] . As such , the ASE ratio can control for the effect of experimental variations on gene expression , which function predominantly in trans [29] , [30] . Here , we used padlock probes and high-throughput sequencing for digital RNA allelotyping to map tissue-specific expression regulatory variants in human iPS cells and their derivatives and showed that allele-specific expression analysis could overcome experimental noise and artifacts . Current approach will allow in vitro experiments on individualized iPS cell lines to map additional tissue-specific and context-dependent regulatory variants .
The Personal Genome Project ( PGP ) is a repository for pre-consented phenotype and genetic data as well as cell lines , including iPS cells . We derived primary skin fibroblast lines from two participants in the Personal Genome Project ( PGP ) , using two partial depth skin biopsy samples obtained from both arms ( Bx1 and Bx2 ) . Clonal populations of PGP1 and PGP9 primary skin fibroblasts ( named PGP1Bx1F and PGP9Bx1F ) were isolated by routine subcloning . Non-clonal populations of primary fibroblasts ( named PGP1Bx2F and PGP9Bx2F ) were derived from a second biopsy site . The PGP1 and PGP9 fibroblast populations were transduced with retrovirus expressing pluripotency reprogramming factors ( OCT4 , SOX2 , KLF4 and MYC ) [31] . The isolated iPS clones expressed pluripotency markers ( Figure 1A ) and formed tetratomas containing normal derivatives of all three germ layers ( Figure 1B ) , confirming their functional pluripotency . In order to harness the accuracy of high-throughput sequencing for quantitative allele-specific RNA analysis , we designed padlock probes targeting 27 , 000 common exonic SNPs ( minor allele frequency > 0 . 07 ) , representing 10 , 345 unique genes , based on the hg18 annotation ( UCSC Genome Browser ) ( Table S1 ) . The padlock probes were synthesized on an Agilent array in a massively parallel manner , and they were then PCR amplified and processed to generate single-stranded DNA molecules [19] , [32] . The pool of single-stranded padlock probes was annealed to the double-stranded cDNA and/or the genomic DNA , followed by a 9-bp fill-in and ligation reaction to circularize the annealed probes [33] , [34] . The circularized products containing the captured sequence were amplified and sequenced on Illumina GAII . On average , we obtained 6 . 4±2 . 0 million sequencing reads per sample , and we were able to map 69 . 8±17 . 2% of the reads against the RefSeq sequences used for the padlock probe design ( Table 1 ) . Approximately 19 , 000 ( 70 . 4% ) out of 27 , 000 SNPs were covered at least 20 times with a mean coverage of 250 reads for each SNP , of which 25% were heterozygous calls ( Table 2 ) . Genotyping calls made using Affymetrix 500K and digital allelotyping showed a concordance rate of 98% for >20x coverage and 99% for >50x coverage ( Table 3 ) . Among the heterozygous SNPs , the ratio between reference and alternative alleles was symmetrically distributed around 0 . 51 ( Figure 2A ) , and the distribution of sequencing reads was nearly identical between the two alleles ( Figure 2B ) , suggesting little or no bias in capturing and mapping the reads . For RNA allelotyping , we amplified the singled stranded cDNA from 50 ng total RNA using linear displacement amplification ( NuGen ) and generated the double stranded cDNA using random hexamer priming ( Invitrogen ) . We confirmed that the padlock probes captured both + and - strands with a similar efficiency , 51 . 6% and 48 . 4% respectively ( Table 4 ) . Typically , we observed ∼1 , 300 ( 25% ) heterozygous expression SNPs out of ∼5 , 200 total expression SNPs . As expected , large ASE deviations were associated with SNPs having a small number of reads ( <100 ) , indicating the presence of biological and/or technical noise ( Figure 3A ) . However , the allele-specific expression ratio was highly reproducible between the total RNA replicates ( R2 = 0 . 7994 with <100 reads and R2 = 0 . 905 with >100 reads ) ( Figure 3C and 3D ) . In order to validate our method , we compared digital RNA allelotyping to quantitative Sanger sequencing , which showed a high correlation between the two methods among the 12 heterozygous expression SNPs in PGP1 samples ( R2 = 0 . 825 ) ( Figure 3B ) . We then asked whether the total number of reads for each SNP might reflect the gene expression level . We compared the mean number of sequencing reads from probes targeting the same transcript and normalized the values against the number of sequencing reads from the genomic DNA . We then compared these values against the relative gene expression levels as measured by Illumina BeadChip Human Ref-8 , revealing only a weak correlation ( R2 = 0 . 1684 ) ( Figure 4A ) . We also asked whether we were capturing only those genes that were highly expressed . When we compared a list of genes captured using our method and compared it to their relative gene expression level , 159 out of 1124 ( 14% ) captured SNPs were associated with the genes below the detection limit on the BeadChip platform ( Figure 4B ) . These results suggested that digital RNA allelotyping was capable of detecting rare transcripts and that the absolute read counts did not necessarily reflect the overall gene expression level , possibly due to differences in probe hybridization , abundance and/or amplification . In our previous study , we showed that human fibroblasts , lymphoblastoid cell lines and primary keratinocytes all demonstrated tissue-specific ASE ( 4 . 3–8 . 5% of heterozygous SNPs ) , using a different probe library design ( CES22k-3 . 2 ) [19] . When adjusted for the false discovery rate in biological replicates , the percentage of SNPs with tissue-specific ASE was between 2 . 3–6 . 5% . Using a new probe design ( CES27k-9bpV3 ) , we looked for tissue-specific ASE in PGP1 fibroblasts and lymphoblastoid cell lines ( Dataset S1 and Dataset S2 ) . We observed that 3 . 8% ( 31/807 ) of the SNPs showed tissue-specific ASE reproducibly in both replicates . Between iPS clones and primary fibroblasts , the number of reproducible tissue-specific ASE loci increased to 9 . 8% ( 107/1091 ) , while it was 6 . 9% ( 71/1036 ) between iPS cells and embryoid bodies ( EBs ) ( Table 5 ) . These findings suggested that up to 10% of ASE showed reproducible tissue-specificity and that they were more numerous in complex and/or heterogeneous tissue samples . In order to explore the relationship of ASE ratios across a wide range of tissue types , we used 186 expression SNPs that were universally present in multiple cell types from PGP1 and 9 and hierarchically clustered them using un-normalized ASE ratios directly ( Figure 5 ) . The sample correlation between the biological replicates was 0 . 983 ( PGP1Bx2 F1 and F2 ) and 0 . 987 ( PGP1Bx1 iPS1a and iPS1c ) , while the correlation between primary fibroblasts and lymphoblastoid cells was 0 . 980 ( PGP1Bx2 fibroblasts versus lymphocytes ) . The differentiated PGP1Bx1 iPS1 derivatives were related to each other with a lower correlation of 0 . 969 . In contrast , the ASE ratio between PGP1 and PGP9 samples had a correlation of only 0 . 542 . We have shown previously that genetic similarity was highly correlated with allelic ratio similarity ( R2 = 0 . 91 ) [19] , and the current result confirmed this conclusion and further suggested that allele-specific expression from human iPS cells were remarkably similar to other cell types from the same individual , despite differences in their epigenetic states [35] . We then normalized direct allelic ratios from the cDNA with those from the genomic DNA in order to reduce probe-specific effects on ASE measurements . To correct for a normalization bias , we calculated the mean ASE ratio across all the samples and used the distance from the mean for hierarchical clustering ( Figure 6 ) . Using the relative change in the ASE ratio across multiple cell types , we observed a consistent correlation between fibroblasts ( 0 . 31 correlation ) , lymphocytes ( 0 . 39 correlation ) and iPS cells ( 0 . 24 correlation ) , while the sample correlation of fibroblasts versus lymphocytes and iPS cells was 0 . 27 and −0 . 0093 , respectively . Finally , the correlation coefficient between the PGP1 and PGP9 samples was −0 . 26 , indicating a significant difference between the two individuals . From these results , we concluded that the structure of cis-regulatory variants was largely genotype-dependent and that the allelic architecture in gene expression changed to a much smaller degree from cell type to cell type . Strictly speaking , the ASE ratio was a quantitative measure that reflected the relative abundance of different RNA alleles . However , any detectable differences in ASE alone could also be used as an indicator of functional regulatory variants nearby . In order to assign a confidence score to ASE-mapped genes , we used a chi-squared test ( cDNA-to-genomic DNA alleles; χ2>6 . 64 ) . Since miniscule ASE could be called ‘significant’ solely due to the large number of sequencing reads , we required that the ASE ratio be >0 . 60 or <0 . 40 . Therefore , our digital ASE calls addressed whether a cis-regulatory variant could be confidently mapped to a gene locus , not whether ASE showed a biologically meaningful allelic imbalance . When examining 427 digital ASE-positive SNPs out of 1822 total SNPs in technical replicates , the correlation coefficient of ASE ratios increased from 0 . 8672 to 0 . 9766 ( Figure 7A ) , suggesting that much of the measurement noise had been eliminated due to a large number of observations . Using technical replicates , we also estimated the false discovery rate of digital ASE calls to be 1 . 6% ( Figure 7B ) , and when all the samples were adjusted for the false discovery rate , 27±4 . 7% of the heterozygous expression SNPs were ‘confidently’ mapped in any given sample ( Table 6 ) . In order to show that digital ASE calls did not depend solely on the number of observations , we compared digital ASE-positive and negative calls and looked at the number of cDNA and genomic DNA reads as well as the average ASE deviation . We observed that the number of cDNA and genomic DNA reads were ∼45% higher , whereas the average allelic ratio deviation was ∼250% higher in the ASE-positive calls ( Table 7 ) . We also examined the ASE calls between PGP1 and PGP9 in order to see if they reflected the difference in allele-specific expression ( Dataset S1 and Dataset S2 ) . While the allelic deviation was ∼90% higher in the ASE-positive calls , the number of genomic DNA reads was also ∼120% higher . These results indicated that our method for mapping ASE-associated regions was influenced by all three parameters , as expected . In order to visualize tissue-specific ASE loci associated with high confidence scores , we examined 1522 heterozygous expression SNPs in 20 PGP1 and PGP9 samples , out of which 317 SNPs were shared among at least 80% of the samples . When these digital ASE calls were hierarchical clustered , there were able to discriminate different tissue types and individuals ( Figure 8A ) . A possible explanation of why digital ASE calls reflected tissue-specificity was that higher tissue-specific expression resulted in higher cDNA observation counts . However , we previously demonstrated that there was no appreciable difference in the number of cDNA reads between ASE-positive and -negative calls in a variety of tissues ( Table 7 ) . In addition , the average number of sequencing reads correlated poorly with the absolute gene expression level ( Figure 4A ) , suggesting that the differences in read counts alone did not explain tissue-specific ASE mapping . We also examined individual-specific ASE-positive clusters with the sample correlation of 0 . 7223 in PGP1 ( 29/317 ) . Interestingly , a large fraction of PGP1-specific clusters were characterized by consistent ASE calls across all cell types ( Figure 8B ) , indicating that approximately 1/3–1/2 of the mapped cis-regulatory variants were cell type and context-independent . So far , we attempted to map cis-regulatory variants using the gene transcripts that were universally present among various cell types and found that up to 10% of the genes might be influenced by tissue-specific regulatory variants . However , we expected that other cis-regulatory variants would only be detected using tissue-specific transcripts . In order to capture these variants , we compared different cell types with a similar sequencing depth ( 5 . 3–7 . 4 million reads ) and counted the number of ASE-positive calls that were specific to that tissue . We were able to examine between 1 , 500 to 1 , 900 heterozygous expression SNPs in primary fibroblasts , immortalized B-lymphocytes , iPS cells and iPS-derived embryoid bodies ( EBs ) from PGP1 ( Table 8 ) . The number of expression SNPs unique to each cell type was 34 ( 2 . 2% ) and 49 ( 3 . 2% ) for fibroblasts and lymphocytes , respectively . In contrast , we observed 126 ( 7 . 8% ) and 287 ( 14 . 9% ) tissue-restricted expression SNPs in iPS cells and EBs , respectively . This suggested that iPS cells and EBs expressed many transcripts absent in primary cell lines . In addition , we found that the percentage of ASE-positive SNPs was generally lower in fibroblast- and lymphocyte-specific transcripts ( ∼24% ) as compared to iPS and EB-specific transcripts ( ∼38% ) . Overall , the number of ASE-positive loci mapped using primary fibroblasts alone was 391 , which increased to 562 ( 44% increase ) using iPS cells and limited in vitro differentiation . We estimated that more than 12% of all heterozygous SNPs were associated with ‘mappable’ functional regulatory variants using our approach . We expect this number to increase when other differentiated cell types are examined . Dosage compensation in mammalian somatic cells is achieved by randomly silencing one of the transcriptionally active X-chromosomes [36] . Random X-inactivation in mouse ES cells is tightly coupled to cell differentiation and the silenced X-chromosome can be re-activated by somatic nuclear transfer [37] . In order to determine how ASE might be affected by re-activation of the silenced X-chromosome after iPS reprogramming , we used a clonal population of female primary fibroblasts to generate two iPS cell lines ( PGP9Bx1 iPS1 and PGP9Bx1 iPS2 ) . We then examined 66 heterozygous expression SNPs that were present on the X-chromosome . We observed 14 genes ( 21% ) that were expressed and captured in the two iPS cell lines from PGP9 . The ASE ratios of these genes were highly reproducible ( R2 = 0 . 98 ) , including 6 out of 14 SNPs ( 42% ) showing a near mono-allelic preference ( Figure 9A ) . We also observed that eight X-chromosomal expression SNPs were shared between PGP9Bx1 F1 and PGP9Bx1 iPS2 . Surprisingly , their ASE ratios were proportionately reversed with a negative linear correlation of R2 = 0 . 52 ( Figure 9B ) . In contrast , the autosomal ASE ratios in the same pair of cell lines demonstrated a positive linear correlation ( R2 = 0 . 63 ) ( Figure 9C ) . When we examined a polyclonal population of primary fibroblasts ( PGP9Bx2F1 ) , their X-chromosomal ASE ratios were near 0 . 5 , likely due to the population averaging of random X-chromosomal inactivation ( Figure 9D ) . These results indicated that both complete and partial inversions of X-chromosomal ASE ratios occurred during iPS reprogramming and that our method was sensitive and robust enough to detect true changes in allele-specific expression due to reasons other than cis-regulatory polymorphisms . We then examined ASE in undifferentiated and differentiating iPS cells . When considering only the ASE-positive SNPs , we observed that the correlation between iPS biological replicates ( R2 = 0 . 94 ) was similar to that of technical replicates ( R2 = 0 . 98 ) ( Figure 10A ) . When iPS cells were treated with 100-µM trans-retinoic acid for 12 hours , the ASE ratio showed a reduction in correlation between replicates ( R2 = 0 . 62 ) , likely due to the heterogeneity of the colony size and the differentiation environment ( Figure 10B ) [38] . When the iPS cells were further differentiated into embryoid bodies ( EBs ) for 7 days , we similarly observed a reduction of correlation between replicates ( R2 = 0 . 59 ) ( Figure 10C ) . We also found that up to 5–13% of the ASE-positive expression SNPs switched the allelic preference during transient and long-term iPS differentiation ( Figure 11A ) , indicating that parental isoforms could be alternately expressed during developmental transitions . While this phenomenon could be due to random stochastic noise , we showed that the ASE ratio was highly reproducible between biological and technical replicates , even among the rare gene transcripts falling below the traditional detection limit . This suggested that ASE switching was due to the biological heterogeneity of stem cell differentiation and not random measurement noise alone . Finally , changes in autosomal ASE did not affect all chromosomes equally during iPS differentiation ( N = 6 samples ) . We observed that Chromosome 6 displayed lower ASE variance that was statistically significant ( p-value: 0 . 022 ) , possibly due to the amount of stable gene imprinting present on Chromosome 6 ( Figure 11B ) . This observation also supported the idea that the variability in ASE during iPS differentiation was not solely due to random noise .
Studying subtle and/or normal variations in gene regulation requires a sensitive and robust method for measuring true genetic effects . Such effects should be measured in a wide range of human tissues , whether by using human tissue samples or in vitro cell culture , both of which can introduce many confounding factors and experimental artifacts . By combining alleles-specific expression analysis together with human pluripotent stem cell reprogramming , we were able to achieve both objectives with high sensitivity and reproducibility . Despite extreme variations in the cell types , the epigenetic status , cell derivation and reprogramming methods and cell differentiation protocols , we were able to detect a subtle allelic imbalance as small as 60:40 and map approximately 27% of the expression SNPs in a given cell line , of which 3–10% were tissue-specific . We also demonstrated that 1/3–1/2 of mappable ASE loci were reproducible regardless of the cell type used and that they were strongly dependent upon the genotype . We also showed that differentiated iPS cells expressed >40% more transcripts associated with ASE and that more should now be mappable using directed in vitro differentiation . Finally , xwe demonstrated two examples of dramatic ASE changes during X-chromosomal inactivation and during iPS differentiation , showing that our approach can successfully detect global changes in allele-specific gene regulation during development . The reproducibility of ASE loci across many different cell types was reassuring , but it also pointed to the possibility of having a systematic bias throughout all the samples . Thus , we asked whether we could find an example of ASE changes that was both expected and biologically interpretable . We found that the X-linked ASE ratio was proportionately inversed after iPS reprogramming , including those that were partially silenced . It was known that up to 25% of the X-linked genes could escape X inactivation in human cell lines [39] , and indeed , we observed 7/23 and 4/16 X-linked SNPs that were only partially silenced in PGP9 iPS cells and fibroblasts , respectively . Our study demonstrated that these genes were still influenced by X inactivation and that the effect remained proportionately similar even after random chromosomal silencing . While nuclear reprogramming has been reported to reset random X-inactivation in cloned mouse embryos [37] and in mouse iPS cells [40] , it was not known whether human iPS cells reached an embryonic ground state . However , we showed that human iPS cells from clonal primary fibroblasts possessed an inverted X-chromosome inactivation pattern , suggesting that human iPS reprogramming can indeed completely erase the somatic X-inactivation memory , a property associated with the embryonic ground state . Conceptually , allele-specific expression is a direct result of functional cis-regulatory mutations or variations . However , it is also caused by random stochastic events [41] , [42] and gene imprinting/silencing [43] as well as allele-specific methylation [22] . Because iPS reprogramming is accompanied by a high degree of cell clonality and epigenetic changes , it offered us an unprecedented opportunity to study how allele-specific expression was affected by such factors . Using a genome-wide allele-specific expression analysis on multiple cell types derived from the same individual , our study conclusively showed that the mappable ASE loci were not dramatically affected by the cell clonality , the methylation status and/or the pluripotency reprogramming and that they were highly individual-specific . It indicated that allele-specific expression might be a good surrogate for indicating the presence of functional cis-regulatory variants . The next logical step will be to determine whether this mappable ASE loci are in fact inheritable and that they can combine in the offspring to produce a gene expression phenotype that is much more dramatic and biologically significant . While it is tempting to use the ASE ratio as a quantitative trait for association mapping , most ASE loci may not produce a strong phenotype in heterozygous individuals . However , allele-specific expression may exert a more direct influence when combined with other functional variants to generate a mixture of functionally altered protein isoforms . With full diploid genome sequencing , it may now be possible to measure the frequency of allelic combinations that may produce measurable effects on the protein function as well as the signaling and/or transcriptional pathways in an allele-specific manner . Our study showed that as many as 5–13% of the mapped ASE loci changed their preference of parent-specific gene expression during early iPS differentiation and development . It will be fascinating to examine whether alternating patterns of parent-specific gene expression associated with functional coding variants can give arise to subtle variations in parent-specific cellular and tissue organization during different phases of the human development . While the most straightforward cause of allele-specific expression is differential transcription factor binding on the promoter , other mechanisms such as alternative splicing and methylation-mediated repression may also play an important role . We are currently developing technologies for examining additional molecular features beyond gene transcription to explore allele-specific processes during gene expression and processing . While functional haploid cells and organisms have greatly enhanced our understanding of various molecular pathways in simple organisms , especially in conjunction with mutagensis screening , such approaches are not possible in higher eukaryotes such as mice and humans . However , an allele-specific readout such as ASE allows one to study the effect of haploid elements and variations in fully functional cell lines , enabling one to design experiments to dissect the phenotypic consequence using family of cell lines with different genetic combinations . Therefore , the real power of ASE and other analyses may not necessarily reside in their ability to map of regulatory variants , but to determine the mechanism of allelic combinations that can contribute to the development of a complex inheritable phenotype . While the use of iPS cells and allele-specific expression analysis for expression trait mapping shows much promise , there are limitations to this approach . The iPS reprogramming , and the propagation and differentiation of iPS cells can be laborious and do not scale up easily . It also does not distinguish among various possible mechanisms for allele-specific expression ( i . e . promoter activation , alternative splicing , sequence-specific degradation ) . In order to bypass these bottlenecks , we are engaged in an effort to automate cell immortalization/iPS reprogramming as well as allele-specific expression assays in order to examine a large population of human volunteers with extensive phenotype and genotype data ( Personal Genome Project ) . Leveraging the power of full genome sequencing technology , our approach of using padlock probes will enable one to examine thousands of samples simultaneously , providing a way to explore cis-regulatory variants in many different tissues in thousands of living study volunteers cost-effectively . We are currently also targeting potential regulatory variants using zinc finger nuclease-mediated homologous recombination in iPS cells to alter their ASE profile and the gene expression level . This and other similar efforts to map and understand numerous functional variants in the vast stretches in the non-coding region and integrating it with experimental biology in a high-throughput manner will likely yield a potent insight into the person-specific regulation in gene expression , cellular biology and ultimately , personalized medicine .
Personal Genome Project ( PGP ) obtained informed consent from human volunteers who have agreed to release both genetic and tissue samples to the research community . All protocols relating to the collection and processing of human data and samples have been approved by Harvard Institutional Review Board ( IRB ) . The primary fibroblasts were maintained in 15% NCS ( Hyclone ) D-MEM/F12 ( Gibco ) supplemented with 10 ng/ml hEGF ( R&D Systems ) , non-essential amino acid ( Gibco ) , Pen/Strep and L-Glutamine ( Gibco ) . The iPS cells were maintained in 20% KO-Serum ( Invitrogen ) KO-DMEM ( Invitrogen ) supplemented with 4 ng/ml bFGF ( BD Biosciences ) , β-ME ( Gibco ) , non-essential amino acid , Pen/Strep and L-Glutamine on a γ-irradiated MEF layer ( GlobalStem ) . Briefly , pMIG containing OCT4 , SOX2 , KLF4 and MYC along with VSV-G and Gag-Pol vectors were transiently transfected into 293T cells . We collected retrovirus-containing medium and passed through a 0 . 45-micron filter unit , followed by ultracentrifugation . We added each virus at multiplicity of infection ( MOI ) of 5 to human primary fibroblasts ( passage number <8 ) . We found that clonally derived PGP1Bx1 fibroblasts were more difficult to reprogram , and it required SV40 large T and NANOG to achieve functional pluripotency [13] . By day 21–30 post-infection , hES cell-like flat colonies started to appear , and they were picked manually and propagated on a freshly prepared MEF layer . The total RNA was prepared using RNeasy ( Qiagen ) . The RNA sample was then linearly amplified and synthesized into a single-strand cDNA using a whole transcriptome amplification method ( NuGen ) . The linearly amplified single-stranded cDNA is then converted into double-stranded cDNA fragments using random hexamers and E . coli DNA polymerase at 16°C for 2 . 5 hours . Of note , we did not observe a significant difference in read counts between the first strand and the second strand ( Table 4 ) . Circularization was performed in 20-ul reactions containing 400 ng genomic DNA or 200 ng ds-cDNA , 0 . 5 pmole padlock probes ( total concentration ) , 2U AmpLigase ( Epicenter ) , 2U AmpliTaq Stoffel fragment ( Applied Biosystems ) , 0 . 1 µM dNTP in 1x AmpLigase buffer . The reactions were incubated at 95°C for 5 minutes , 60°C for 48 hours . The reactions were then denatured at 94°C for 1 minutes , cooled down to 37°C , then digested with Exonuclease I ( 10U ) and Exonuclease III ( 100U ) for 2 hours at 37°C , and finally heat inactivated at 94°C for 5 minutes . Post-capturing PCR reactions were performed in 100-ul reactions including 10-ul circularization products , 0 . 4x SYBR Green I , 0 . 4 µM forward and reverse PCR primers in 1x iProof PCR master mix . The parameter for real-time PCR was 98°C 30 seconds; followed by 3 cycles of 98°C 15 seconds , 53°C 20 seconds , 72°C 10 seconds; then <15 cycles of 98°C 15 seconds and 72°C 20 seconds . We terminated the reactions when the amplification curves went up close to the plateau stage . The 154-bp amplicon was purified with a 6% TBE polyacrylamide gel ( Invitrogen ) , and sequenced with Illumina Genome Analyzer II . We designed the padlock probes to ensure that the captured sequences are uniquely mappable to the genome using UCSC BLAT . We mapped sequencing reads ( 25–41 bp ) to the sequences by NCBI BLAST using the word size of 8–12 depending on the read length , considering the variant site as degenerate ( NCBI Short Read Archive #SRA008291 . 1 ) . For any sequences that had more than one hit , we required that the second hit had an e-value 5-fold higher than the top hit . In contrast , Maq-based mapping could not handle degenerate sequences , and it was consistently biased towards the reference allele . We made genotyping calls using the “best-P” method on SNPs that were sampled at least 20 times . For each SNP we performed both the test of homozygosity ( assuming the allelic ratio of ( 1-e ) /e where e is the sequencing error ) and the test of heterozygosity ( assuming 50:50 allelic ratio ) , and determined the genotype based on the one that giving a higher p-value . We used chi-squared test to identify expressed SNPs that exhibit RNA allelic ratios significantly different from the genomic allelic ratios ( see Table S1 , Dataset S1 , Dataset S2 ) . Hierarchical clustering and image viewing were done on Cluster and TreeView . | Most complex traits likely result from a combination of genetic polymorphisms . The normal variation in gene expression is thought to be an important contributor . In order to examine a wide range of personalized tissue types from a given individual , we developed a robust method for detecting regulatory variants genome-wide in human induced pluripotent stem ( iPS ) cells . By having a platform capable of mapping regulatory variants despite large biological and experimental noise , and by being able to use in vitro differentiation to derive multiple human tissue types , our approach should enable the identification of large numbers of regulatory variants genome-wide using minimally invasive skin biopsies from a large number of human subjects . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/gene",
"expression",
"genetics",
"and",
"genomics/functional",
"genomics"
] | 2009 | A Robust Approach to Identifying Tissue-Specific Gene Expression Regulatory Variants Using Personalized Human Induced Pluripotent Stem Cells |
Activation of NLRP3 inflammasome is important for effective host defense against invading pathogen . Together with apoptosis-associated speck-like protein containing CARD domain ( ASC ) , NLRP3 induces the cleavage of caspase-1 to facilitate the maturation of interleukin-1beta ( IL-1β ) , an important pro-inflammatory cytokine . IL-1β subsequently plays critical roles in inflammatory responses by activating immune cells and inducing many secondary pro-inflammatory cytokines . Although the role of NLRP3 inflammasome in immune response is well defined , the mechanism underlying its assembly modulated by pathogen infection remains largely unknown . Here , we identified a novel mechanism by which enterovirus 71 ( EV71 ) facilitates the assembly of NLRP3 inflammasome . Our results show that EV71 induces production and secretion of IL-1β in macrophages and peripheral blood mononuclear cells ( PBMCs ) through activation of NLRP3 inflammasome . EV71 replication and protein synthesis are required for NLRP3-mediated activation of IL-1β . Interestingly , EV71 3D protein , a RNA-dependent RNA polymerase ( RdRp ) was found to stimulate the activation of NLRP3 inflammasome , the cleavage of pro-caspase-1 , and the release of IL-1β through direct binding to NLRP3 . More importantly , 3D interacts with NLRP3 to facilitate the assembly of inflammasome complex by forming a 3D-NLRP3-ASC ring-like structure , resulting in the activation of IL-1β . These findings demonstrate a new role of 3D as an important player in the activation of inflammatory response , and identify a novel mechanism underlying the modulation of inflammasome assembly and function induced by pathogen invasion .
The innate immune system is a highly conserved signaling network important for protection of the infected host and clearance of the invading pathogen [1] . Recognition of the pathogen-associate molecular patterns ( PAMPS ) is dependent on host pattern recognition receptors ( PRRs ) , whose activation results in the production of interferons ( IFNs ) and pro-inflammatory cytokines . Several families of PRRs have been identified , including the Toll-like receptor ( TLR ) [2] , the RIG-I-like receptor ( RLR ) [3] , the NOD-like receptor ( NLR ) [4] , and the C-type lectin receptor ( CLR ) [5] . An important part of the innate immune response is the activation of inflammasome , a cytosolic complex of proteins that activates caspase-1 ( Casp-1 ) to produce the pro-inflammatory cytokine interleukin-1beta ( IL-1β ) [6] . One of the best-characterized inflammasomes consists of NLR family PYRIN domain containing-3 ( NLRP3 ) that harbors an N-terminal PYRIN domain ( PYD ) , a NACHT-associated domain ( NAD ) , and a C-terminal leucine-rich repeat ( LRR ) [7] . The PYD domain of NLRP3 interacts with the PYD domain of the adaptor protein , apoptosis-associated speck-like protein with CARD domain ( ASC ) . NLRP3 oligomerizes through homotypic interactions between NACHT domains following the detection of pathogen infection or cellular stress . The LRR domain has been implicated in ligand sensing and auto-regulation . NLRP3 inflammasome is activated upon exposure to pathogens , including bacteria ( Listeria monocytogenes and Staphylococcus aureus ) [8] and viruses ( Sendai virus , adenovirus and influenza virus ) [9] [10] , by host-derived molecules , such as extracellular glucose [11] , extracellular ATP [12] , and hyaluronan [13] , and also detects signs of metabolic stress and environmental irritants [7] . Together with ASC , NLRP3 promotes the cleavage of pro-Casp-1 to generate active subunits p20 and p10 , which regulate the maturation of IL-1β [14] . IL-1β plays an important role in inflammatory response by recruitment and activation of immune cells as well as production of secondary pro-inflammatory cytokines [15] . NLRP3 can recognize many RNA viruses to regulate innate immunity and viral replication , including enterovirus 71 ( EV71 ) [16] . EV71 is a highly infectious RNA virus causing hand-foot-mouth disease ( HFMD ) , meningoencephalitis , neonatal sepsis , and even fatal encephalitis in children [17] . EV71 induces many pro-inflammatory cytokines that play important roles in the development of inflammation and associated diseases [18] . Although NLRP3 inflammasome plays important role in regulating host innate immune response and viral infection , the assembly and activation of NLRP3 inflammasome mediated by viral infection are poorly understood . In this study , we identify a novel mechanism by which EV71 facilitates the assembly of NLRP3 inflammasome . EV71 was found to stimulate the cleavage of pro-Casp-1 and the secretion of IL-1β by modulating the components of NLRP3 inflammasome . More significantly , EV71 3D protein was shown to play a stimulatory role in the activation of NLRP3 inflammasome . 3D protein is a RNA-dependent RNA polymerase ( RdRp ) essential for viral replication [19] . It activates the inflammasome through direct binding to NLRP3 . The interaction of 3D with NLRP3 facilitates the assembly of inflammasome complex by forming a “3D-NLRP3-ASC” ring-like structure . These findings demonstrate a novel mechanism underlying the regulation of inflammasome complex assembly in response to pathogen infection , which would provide insights into the prevention and treatment of the viral infection .
EV71 infection induces pro-inflammatory cytokines that play important roles in the development of inflammation and associated diseases [20] . Among the pro-inflammatory cytokines , IL-1β plays important roles in the induction of inflammation by recruitment and activation of immune cells and production of secondary pro-inflammatory cytokines [21] . Thus , we determined the effect of EV71 on the production and secretion of IL-1β . TPA-differentiated THP-1 macrophages were infected with EV71 . The secretion of IL-1β ( Fig 1A ) and cleavage of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) ( Fig 1B , upper panel ) in supernatants , and the production of pro-IL-1β in the cell lysates ( Fig 1B , lower panel , lanes 3–4 ) were induced by EV71 infection . EV71 3D protein was expressed and VP1 mRNA was detected ( S1A Fig ) during EV71 infection ( Fig 1B , lanes 3–4 ) , indicating that EV71 replicated well in the cells . In addition , THP-1 macrophages were treated with lipopolysaccharides ( LPS ) and Nigericin ( a bacterial toxin that activates NLRP3 by causing potassium efflux in a pannexin-1-dependent pathway ) positive control . The expression of IL-1β mRNA ( Fig 1C ) was stimulated by LPS+Nigericin and EV71 . It is clear that EV71 induces IL-β in THP-1 macrophage . Next , TPA-differentiated THP-1 macrophages were treated with LPS and Nigericin , or infected with EV71 at different multiplicity of infections ( MOI ) , as indicated . The secretion of IL-1β ( Fig 1D ) and cleavage of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) ( Fig 1E , top panel ) in the cell supernatants , and the production of pro-IL-1β in the cell lysates ( Fig 1E , bottom panel , lanes 3–5 ) were induced by EV71 in dose-dependent manners . Similarly , the expression of IL-1β mRNA was activated by LPS/Nigericin and EV71 virus ( Fig 1F ) . EV71 3D protein ( Fig 1E ) and VP1 mRNA ( S1B Fig ) were detected in viral infected cells , suggesting that EV71 replicated efficiently in the macrophages . Therefore , these results indicated that EV71 induces the production and secretion of IL-1β in dose-dependent manners . The activation of IL-1β is regulated by two pathways in response to pathogen infection: transcription of pre-IL-1β mRNA and proteolytical processing of pre-IL-1β protein by caspase-1 [7] . Previous study has demonstrated that EV71 infection activates NF-κB in rat vascular smooth muscle cells and SK-N-SH cells [22 , 23] . We also demonstrated that the endogenous genes regulated by NF-κB were activated by EV71 and by lipopolysaccharide ( LPS ) , an NF-κB-activating stimuli ( S1C Fig ) . ASC oligomers were formed by EV71 infection and Nigericin treatment , indicating that EV71 and Nigericin induce inflammasome activation [24] ( S1D Fig ) . Taken together , our results demonstrate that the production and secretion of IL-1β are induced by EV71 in macrophages . The mechanism by which EV71 activates IL-1β was investigated . As maturation of IL-1β is mediated by the NLRP3 inflammasome [6] , EV71 may activate IL-1β through modulating the NLRP3 inflammasome . To determine this possibility , HEK293T cells were co-transfected with plasmids encoding the components of NLRP3 complex ( NLRP3 , ASC , and pro-caspsase-1 ) along with the substrate of NLRP3 inflammasome ( pro-IL-1β ) , and then treated with ATP and infected with EV71 . The secretion of IL-1β ( p17 ) was up-regulated by ATP and EV71 in cell supernatants ( Fig 2A ) , and the production of IL-1β ( p17 ) and Casp-1 subunit ( p20 ) were enhanced by ATP or EV71 in cell lysates ( Fig 2B ) . These results suggest that EV71 facilitates the activation of IL-1β and Casp-1 in the presence of NLRP3 inflammasome components ( Casp-1 , NLRP3 , and ASC ) in HEK293T cells . The roles of components of NLRP3 inflammasome complex in the production and secretion of IL-1β were further evaluated by short hairpin RNAs ( shRNAs ) mediated knockdown of gene expression . Four stable TPA-differentiated THP-1 macrophages cell lines were generated , which stably expressed short hairpin RNAs ( shRNAs ) targeting NLRP3 ( sh-NLRP3 ) , ASC ( sh-ASC ) , and pro-caspase-1 ( sh-Casp-1 ) , respectively . In sh-NLRP3 stable cells , NLRP3 mRNA was down-regulated , whereas ASC and pro-Casp-1 mRNAs were not affected ( S2A Fig ) . In sh-ASC stable cells , ASC mRNA was attenuated , but NLRP3 and pro-Casp-1 mRNAs remain the same ( S2B Fig ) . In sh-Casp-1 stable cells , pro-Casp-1 mRNA was reduced , while NLRP3 and ASC mRNAs were unchanged ( S2C Fig ) . In addition , NLRP3 , ASC , and pro-Casp-1 proteins were attenuated in sh-NLRP3 stable cells , sh-ASC stable cells , and sh-Casp-1 stable cells , respectively ( S2D Fig ) . These results indicated that shRNAs were stably expressed and specifically silenced their target gene expression . The stable cell lines were then treated with Nigericin . In the cell supernatants , secretion of IL-1β was induced by Nigericin , but its activation was significantly attenuated by sh-NLRP3 , sh-ASC , and sh-Casp-1 , respectively ( Fig 2C ) . Similarly , the cleavage of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) were enhanced by Nigericin ( Fig 2D , top panel , lane 5 ) , but their levels were significantly down-regulated by sh-NLRP3 , sh-ASC , and sh-Casp-1 , respectively ( Fig 2D , top panel , lanes 6–8 ) . These results demonstrate that shRNA-mediated gene silence of the components ( NLRP3 , ASC , or pro-caspase-1 ) of NLRP3 inflammasome attenuates the production and secretion of IL-1β . The role of NLRP3 inflammasome components in EV71-induced activation of IL-1β was further explored in the stable cell lines infected with EV71 . In the cell supernatants , the secretion of IL-1β ( p17 ) was increased by EV71 , but significantly attenuated by sh-NLRP3 , sh-ASC , and sh-Casp-1 , respectively ( Fig 2E ) . The cleavage of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) was enhanced by EV71 ( Fig 2F , top panel , lane 5 ) , but significantly reduced by sh-NLRP3 , sh-ASC , or sh-Casp-1 ( Fig 2F , top panel , lanes 6–8 ) . In the cell lysates , the productions of matured IL-1β ( p17 ) were stimulated by EV71 ( Fig 2F , bottom panel , lane 5 ) , but down-regulated by sh-NLRP3 , sh-ASC , or sh-Casp-1 , respectively ( Fig 2F , bottom panel , lane 6–8 ) . These results suggest that activation of IL-1β by EV71 requires the components of NLRP3 inflammasome complex . Taken together , our results demonstrate that EV71 activates IL-1β through regulating the NLRP3 inflammasome . The mechanism involved in the modulation of NLRP3 inflammasome mediated by EV71 was investigated . The effect of EV71 replication on the activation of IL-1β was evaluated . TPA-differentiated THP-1 macrophages were treated with LPS/Nigericin , incubated with ultraviolet ( UV ) -inactivated EV71 or heat-inactivated EV71 , or infected with infectious EV71 , as indicated . In the cell supernatants , the secretion of IL-1β was significantly increased by LPS/Nigericin and infectious EV71 and modestly up-regulated by UV- and heat-inactivated EV71 ( Fig 3A ) . The cleavage of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) were significantly enhanced by LPS/Nigericin and infectious EV71 , but not affected by UV- and heat-inactivated EV71 in the cell supernatants ( Fig 3B , top panel ) . EV71 3D was expressed only in the cells infected with infectious EV71 ( Fig 3B , bottom panel , lane 5 ) , but not detected in the cells inoculated with UV- and heat-inactivated EV71 ( Fig 3B , lanes 3–4 ) , suggesting that UV- and heat-inactivated EV71 failed to replicate in the cells . Therefore , these results indicate that the replication of EV71 is required for the activation of IL-1β in macrophages . Additionally , the effect of EV71 replication on the activation of IL-1β in human primary peripheral blood mononuclear cells ( PBMCs ) was evaluated . PBMCs were infected with EV71 , treated with LPS , and inoculated with UV-inactivated or heat-inactivated EV71 , respectively . EV71 VP1 mRNA was detected in the cells infected with EV71 , but not in the cells treated with LPS , or inoculated with UV-inactivated or heat-inactivated EV71 ( S3A Fig ) , indicating that EV71 replicates in PBMCs , but UV-inactivated EV71 and heat-inactivated EV71 failed to replicate . The secretion of IL-1β was stimulated by LPS , activated by EV71 in an MOI-dependent manner , but not affected by UV-inactivated or heat-inactivated EV71 ( Fig 3C ) . The expression of IL-1β mRNA was stimulated by LPS , activated by EV71 , but not by UV-inactivated or heat-inactivated EV71 ( Fig 3D ) . These results suggested that the replication of EV71 is required for the production and secretion of IL-1β in PBMCs . It is well established that NLRP3 inflammasome mediated innate immunity to virus through recognition of viral RNA [25 , 26] . To further determine whether EV71 RNA alone can initiate the activation of the NLR3 inflammasome , we stimulated THP-1 macrophages with EV71 viral genomic RNA , HCV viral genomic RNA , and poly ( dA:dT ) as a positive control . The results showed that IL-1β secretion was induced by HCV genomic RNA and poly ( dA:dT ) , but not by EV71 genomic RNA in the cell supernatants of THP-1 macrophages ( Fig 3E ) . The cleavage of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) were not detected in the presence of transfected EV71 RNA ( Fig 3F ) . The levels of pro-IL-1β mRNA and TNF-α mRNA were enhanced by the treatment of EV71 viral genomic RNA and HCV viral genomic RNA ( S3B Fig ) . These results demonstrated that , unlike HCV genomic RNA and influenza A virus RNA , EV71 genomic RNA is unable to activate NLRP3 inflammasome . Previous study has demonstrated that rapid NLRP3 inflammasome activation is independent of “priming” , given that both NF-κB activation and new protein synthesis are not necessary [27] . CHX-pretreated cells produced IL-1β normally in response to extracellular ATP which indicated that NLRP3 inflammasome activation by ATP stimulation does not require de novo translation . In contrast , pretreatment of cells with cycloheximide ( CHX ) , a protein synthesis inhibitor , significantly inhibited EMCV-induced IL-1β secretion . This indicated that virus-encoded proteins activate the NLRP3 inflammasome [28] . We further determined whether EV71 viral protein synthesis was also required for the activation of IL-1β . TPA-differentiated THP-1 macrophages were infected with EV71 for 2 h , treated with CHX for 1 h , and then grown for additional 22 h . In the absence of CHX , the secretion of IL-1β was stimulated by Nigericin treatment ( Fig 4A , lane 4 ) or EV71 infection ( Fig 4A , lane 7 ) . In the presence of CHX , Nigericin-induced secretion of IL-1β was relatively unaffected ( Fig 4A , lane 6 ) , but EV71-activated secretion of IL-1β was significantly repressed ( Fig 4A , lane 9 ) . In the absence of CHX , the cleavages of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) were stimulated by Nigericin treatment ( Fig 4B , lane 4 ) or by EV71 infection ( Fig 4B , lane 7 ) in the cell supernatants . In the presence of CHX , Nigericin-activated cleavages of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) were relatively unaffected ( Fig 4B , lane 6 ) , whereas EV71-activated cleavages of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) were significantly reduced ( Fig 4B , lane 9 ) . The production of EV71 3D ( Fig 4B , lanes 7–9 ) was significantly attenuated in the presence of CHX ( Fig 4B , lane 9 ) . Therefore , the activation of IL-1β mediated by Nigericin do not require de novo protein synthesis , whereas the activation of IL-1β induced by EV71 requires de novo protein synthesis . Taken together , we demonstrated that EV71 replication and protein synthesis are essential for the activation of IL-1β in macrophages and PBMCs . The roles of EV71 nonstructural proteins in the regulation of NLRP3 inflammasome were further determined , as EV71 protein synthesis is required for such regulation . Initially , the activity of NLRP3 inflammasome was determined in HEK293T cells co-transfected together with plasmids encoding NLRP3 , ASC , pro-Casp-1 , and pro-IL-1β . In the cell supernatants , the secretion of IL-1β ( p17 ) was not detected in the presence of NLRP3 alone ( Fig 4C , lane 2 ) or ASC alone ( Fig 4C , lane 3 ) , detected at lower level in the presence of Casp-1 ( Fig 4C , lane 4 ) or NLRP3 plus ASC ( Fig 4C , lane 5 ) , and significantly activated in the presence of all three components , NLRP3 , ASC , and Casp-1 ( Fig 4C , lane 6 ) . In the cell lysates , the levels of pro-IL-1β and Casp-1 ( p22/p20 ) were relatively unchanged in the presence of one or two components ( Fig 4C , lanes 1–5 ) , but significantly stimulated in the presence of all three components ( Fig 4C , lane 6 ) . These results indicated that all components of NLRP3 inflammasome are required for the function of the inflammasome , and suggested that NLRP3 inflammasome is functionally effective under these conditions . HEK293T cells were co-transfected together with plasmids encoding NLRP3 , ASC , pro-Casp-1 , pro-IL-1β , and then transfected with each of the EV71 non-structure proteins , 2A , 2B , 2C , 3A , 3C , and 3D , respectively , as indicated . The secretion of IL-1β in the cell supernatants was inhibited by 2A , relatively unaffected by 2B and 3A , down-regulated by 2C and 3C , and activated by 3D ( Fig 4D ) . Similarly , the production of IL-1β ( p17 ) in the cell lysates was inhibited by 2A ( Fig 4E , lane 2 ) , unaffected by 2B and 3A ( Fig 4E , lanes 3 and 5 ) , enhanced by 2C ( Fig 4E , lane 4 ) , reduced by 3C ( Fig 4E , lane 6 ) , but activated by 3D ( Fig 4E , lane 7 ) . In addition , the cleavage of caspsase-1 ( p20 ) was inhibited by 2A ( Fig 4E , lane 2 ) , unaffected by 2B and 2C ( Fig 4E , lanes 3 and 4 ) , reduced by 3A and 3C ( Fig 4E , lanes 5 and 6 ) , but significantly stimulated by 3D ( Fig 4E , lane 7 ) . Moreover , NLRP3 was inhibited by 2A ( Fig 4E , lane 2 ) , unaffected by 2B , 2C , and 3A ( Fig 4E , lanes 3–5 ) , repressed by 3C ( Fig 4E , lane 6 ) , and enhanced by 3D ( Fig 4E , lane 7 ) . These results revealed that EV71 2A and 3C repress the production and secretion of IL-1β through attenuating NLRP3 inflammasome , which is consistent with a previous report [16] . More interestingly , our results demonstrated that EV71 3D was the only viral protein that stimulates the secretion of IL-1β ( Fig 4D and 4E ) and the cleavage of pro-Casp-1 ( Fig 4E ) . The correlation between the opposite functions of EV71 proteins was then evaluated in TPA-differentiated THP-1 cells infected with EV71 at MOI = 20 for different time . The results showed that the secretion of IL-1β was induced by EV71 at 6 h p . i . ( Fig 4F , lane 3 ) and then increased as the infection time increased ( Fig 4F , lanes 4–6 ) . The cleavages of IL-1β ( p17 ) and caspase-1 ( p20 and p22 ) were stimulated by EV71 at 6 h p . i . ( Fig 4G , lane 3 ) . Similarly , EV71 3D protein was also detected at 6 h p . i . ( Fig 4G , lane 3 ) and then increased as the infection time increased ( Fig 4G , lanes 4–6 ) . However , EV71 3C protein was detected at 12 h p . i . ( Fig 4G , lane 4 ) and then increased as the infection time increased ( Fig 4G , lanes 5 and 6 ) . EV71 2A protein was not detect by western blot , likely due to the concomitant restriction on its expression from its inhibition effect on host gene expression [29] . Thus , EIF4G , a known substrate of 2A , was used as an indictor for the activity of 2A proteases . The results showed that EIF4G protein was reduced at 36 h p . i . ( Fig 4G , lane 6 ) . The temporal expression of EV71 2A , 3C , and 3D were further confirmed in RD cells infected with EV71 at different times . The results showed that 3D expression was initiated at 6 h p . i . ( Fig 4H , lane 4 ) , whereas EIF4G cleavage and 3C expression was started at 12 h p . i . ( Fig 4H , lane 5 ) . Taken together , these results revealed that the secretion of IL-1β or the activation of NLRP3 inflammasome was stimulated by EV71 3D and that 3D protein can overcome the inhibitory effects of EV71 2A and 3C proteins in the secretion of IL-1β . Therefore , we focused the rest of study on determining the role of EV71 3D in the activation of NLRP3 inflammasome and the mechanism underlying such regulation . EV71 3D induces the cleavage of pro-Casp-1 and the secretion of IL-1β through suggesting that it may activate NLRP3 inflammasome , as the NLRP3 inflammasome is critical for pro-Casp-1 activation and pro-IL-1β procession . The effect of EV71 3D on the activation of NLRP3 inflammasome was determined . HEK293T cells were co-transfected together with plasmids encoding NLRP3 , ASC , pro-Casp-1 , and pro-IL-1β , along with plasmid encoding EV71 3D , as indicated . The productions of IL-1β ( p17 ) and Casp-1 ( p22/p20 ) were not detected in the presence of one or two components of the inflammasome ( Fig 5A , lanes 1–3 and 5–7 ) , activated in the presence of NLRP3 , ASC , and pro-Casp-1 ( Fig 5A , lane 4 ) , and further facilitated by 3D ( Fig 5A , lane 8 ) . Similarly , the secretion of IL-1β was not detected in the presence of one or two components of the inflammasome ( Fig 5B , lanes 1–3 and 5–7 ) , stimulated in the presence of NLRP3 , ASC , and pro-Casp-1 ( Fig 5B , lane 4 ) , and further enhanced by 3D ( Fig 5B , lane 8 ) . These results demonstrated that EV71 3D enhances the activity of NLRP3 inflammasome to facilitate the production and release of IL-1β . In addition , the role of EV71 3D in regulating the activity of NLRP3 inflammasome was evaluated in TPA-differentiated THP-1 macrophages which infected with lentivirus expressing EV71 3D protein for the stable protein expression cell lines . The results showed that 3D activated the secretion of IL-1β ( p17 ) in the cell supernatants ( Fig 5C , top panel ) , as well as the production of IL-1β ( p17 ) and Casp-1 ( p22 ) in the cell lysates ( Fig 5C , bottom ) . Furthermore , the effect of EV71 3D on ASC pyroptosome formation was evaluated in TPA-differentiated THP-1 macrophages infected with lentiviruses expressing 3D . We used the TPA-differentiated THP-1 macrophages as a negative control and the TPA-differentiated THP-1 macrophages which were treated with Nigericin as a positive control . ASC pyroptosome formation was activated by 3D and Nigericin , respectively ( Fig 5D ) . In addition , the role of EV71 in regulating the activity of NLRP3 inflammasome was evaluated in TPA-differentiated THP-1 macrophages infected with lentiviruses expressing EV71 3D , treated with Nigericin or ATP , and infected with Sendai virus ( SeV ) . The secretion of IL-1β ( p17 ) was activated by Nigericin , ATP , and SeV in the cell supernatants , whereas such activations were significantly facilitated by 3D ( Fig 5E ) . Similarly , the secretion of IL-1β ( p17 ) was stimulated by Nigericin , ATP and SeV , whereas 3D further facilitated the secretion of IL-1β ( p17 ) mediated by Nigericin , ATP , and SeV ( Fig 5F ) . Previous study has demonstrated that aurintricarboxylic acid ( ATA ) was able to inhibit the RdRp activity of EV71 3D protein , but not inhibit the protease activities of 2A and 3C [30] . Our results revealed that lentivirus expressed EV71 3D activated the secretion of IL-1β in THP-1 cells in the absence of ATA ( Fig 5G , lane 2 ) , but failed to induce the secretion of IL-1β in the presence of ATA ( Fig 5G , lane 4 ) . In the absence of ATA , the secretion of IL-1β ( p17 ) was activated by ATP ( Fig 5G , lane 5 ) and such activation was significantly facilitated by 3D ( Fig 5G , lane 6 ) . In the presence of ATA , the secretion of IL-1β ( p17 ) was also activated by ATP ( Fig 5G , lane 7 ) , but EV71 3D had no affect on ATP-induced secretion of IL-1β ( Fig 5G , lane 8 ) . Similarly , EV71 3D protein stimulated the cleavage of IL-1β ( p17 ) in the absence of ATA ( Fig 5H , lane 2 ) , but failed to induced the cleavage of IL-1β ( p17 ) in the presence of ATA ( Fig 5H , lane 4 ) . In the presence of ATA , the cleavage of IL-1β was induced by ATP ( Fig 5H , lane 7 ) , but EV71 3D failed to facilitate ATP-induced secretion of IL-1β ( Fig 5H , lane 8 ) . In addition , 3D protein production was not affected by ATA ( Fig 5H , lane 2 vs 4 , and lane 6 vs 8 ) , suggesting that the RdRp activity of EV71 3D may be required for the activation of NLRP3 inflammasome . Moreover , TPA-differentiated THP-1 macrophages were infected with EV71 and treated with ATA and ATP . The results showed that ATP-induced secretion of IL-1β ( Fig 5I , lane 3 ) was slightly reduced by ATA ( Fig 5I , lane 3 ) , but EV71-induced secretion of IL-1β ( Fig 5I , lane 5 ) were significant attenuated by ATA ( Fig 5I , lane 6 ) . Similarly , ATP-induced cleavage of IL-1β ( Fig 5J , lane 3 ) was slightly reduced by ATA ( Fig 5J , lane 4 ) , but EV71-induced cleavage of IL-1β ( Fig 5J , lane 5 ) were significant attenuated by ATA ( Fig 5J , lane 6 ) . These results demonstrated that RdRp activity of EV71 3D is essential for NLRP3 inflammasome activation . Taken together , we revealed that EV71 3D protein induces the activity of NLRP3 inflammasome and the productions of IL-1β ( p17 ) and Casp-1 ( p22/p20 ) , and that the RdRp activity of EV71 3D is essential for EV71-induced activation of NLRP3 inflammasome . The mechanism underlying the regulation of NLRP3 inflammasome mediated by EV71 3D was elucidated . Initially , we determined whether EV71 3D is interacted with the components of NLRP3 inflammasome . Interestingly , yeast strain AH109 was co-transformed with the combination of AD-3D and BD-NLRP3 inflammasome components or the three functional domains of NLRP3 . We revealed that EV71 3D was interacted with NLRP3 LRR domain ( S4A and S4B Fig ) . The interaction between EV71 3D and NLRP3 was verified by co-immunoprecipitation ( CoIP ) assays in HEK293T cells co-transfected with plasmid expressing Flag-NLRP3 or HA-3D , as indicated . CoIP results showed that 3D was associated with NLRP3 ( Fig 6A ) and NLRP3 was interacted with 3D ( Fig 6B ) . The interaction between EV71 3D and NLRP3 was further determined in TPA-differentiated THP-1 macrophages infected with lentiviruses expressing 3D . CoIP results indicated that 3D was also interacted with NLRP3 in the treated macrophages ( Fig 6C ) . These results demonstrated that EV71 3D protein can interact with NLRP3 protein . NLRP3 contains several prototypic domains , including PYD , NACHT , and LRR domains . Four plasmids expressing NLRP3 , PYRIN domain , NACHT domain , and LRR domain were constructed ( S4C Fig ) . The domains of NLRP3 involved in the interaction with EV71 3D were then determined . HEK293T cells were co-transfected with plasmid expressing HA-3D along with plasmids expressing Flag-NLRP3 , Flag-PYD , Flag-NACHT , and Flag-LRR , respectively . CoIP results showed that 3D was interacted with NLRP3 ( Fig 6D , lane 2 ) , NACHT domain ( Fig 6D , lane 6 ) , and LRR domain ( Fig 6D , lane 8 ) , but not PYD domain ( Fig 6D , lane 4 ) . Similarly , NLRP3 ( Fig 6E , lane 2 ) , NACHT domain ( Fig 6E , lane 6 ) , and LRR domain ( Fig 6E , lane 8 ) , but not PYD domain ( Fig 6E , lane 4 ) , were interacted with 3D . Therefore , we demonstrated that EV71 3D binds with NLRP3 through interacting with the NACHT and LRR domains . We then determined whether 3D also interacts with other components ( ASC and Casp-1 ) of NLRP3 inflammasome through interacting with NLRP3 . Upon activation , NLRP3 is oligomerized , which leads to NLRP3 PYD domain clustering and presentation for interaction with ASC PYD domain , and ASC CARD domain subsequently recruits pro-Casp-1 CARD domain to permit the auto-cleavage and the formation of active Casp-1 p10/p20 . Thus , we evaluated the ability of 3D in the interaction with ASC . CoIP assays showed that 3D was interacted with NLRP3 ( S5 Fig , lane 2 ) , but not ASC ( S5 Fig , lane 4 ) . The interaction between 3D and ASC was further investigated in HEK293T cells co-transfected with plasmids expressing NLRP3 , PYRIN domain , ASC , and 3D , as indicated . 3D was associated with ASC only in the presence of NLRP3 ( Fig 7A , lane 2 ) , but not in the presence of PYRIN domain ( Fig 7A , lane 4 ) . Similarly , ASC was associated with 3D only in the presence of NLRP3 ( Fig 7B , lane 2 ) , but not in the presence of PYRIN domain ( Fig 7B , lane 4 ) . NLRP3 PYRIN domain interacts with ASC , but not with 3D , suggesting that 3D is associated with ASC through interacting with NLRP3 . In addition , the interaction of 3D with NLRP3 , ASC , and pro-Casp-1 was evaluated by GST pull-down assays . HEK293T cells were co-transfected with plasmid expressing Flag-NLRP3 and plasmids encoding Flag-pro-Casp-1 , Flag-ASC , and GST-3D , as indicated . The results showed that 3D could interact with NLRP3 even in the absence of ASC and pro-Casp-1 ( Fig 7C , lane 2 ) , and could also interact with NLRP3 but not with pro-Casp-1 in the absence of ASC ( Fig 7C , lane 4 ) . However , 3D was associated with NLRP3 , ASC , and pro-Casp-1 in the presence of all components of NLRP3 inflammasome ( Fig 7C , lane 6 ) . The purified GST-LRR could interact with HA-3D protein which also demonstrated the interaction between NLRP3 and EV71 3D protein ( Fig 7D ) . These results indicated that 3D is associated with ASC or pro-Casp-1 through interacting with NLRP3 . Moreover , the interactions of 3D with endogenous NLRP3 and ASC were explored in TPA-differentiated THP-1 macrophages infected with EV71 . The results showed that 3D was co-immunoprecipitated with endogenous NLRP3 and ASC ( Fig 7E ) . Thus , EV71 3D is associated with the components ( NLRP3 , ASC , and pro-Casp-1 ) of NLRP3 inflammasome through interacting with NLRP3 . As 3D interacts with NLRP3 , the effect of 3D on the sub-cellular distribution of NLRP3 was examined under confocal microscopy . We evaluated the effect of EV71 on the sub-cellular distribution of NLRP3 in TPA-differentiated THP-1 macrophages infected with EV71 by immunofluorescence assays . In mock-infected macrophages ( Fig 7Fa to 7Fd ) , NLRP3 was diffusely distributed in the cell cytosol ( Fig 7Fa and 7Fd ) . However , in EV71-infected cells ( Fig 7Fe to 7Fh ) , NLRP3 was co-localized with 3D ( Fig 7Fe and 7Ff ) to form distinct spots in the cell cytosol ( Fig 7Fh ) . These results revealed that 3D interacts with NLRP3 to form a 3D-NLRP3 complex in the cytoplasm during EV71 infection . The role of EV71 on the sub-cellular distribution of ASC was also determined in TPA-differentiated THP-1 macrophages infected with or without EV71 . In mock-infected macrophages ( S6a to S6c Fig ) , ASC was diffusely distributed in the nucleus and cytosol ( S6a and S6c Fig ) . In EV71-infected cells ( S6d to S6f Fig ) , ASC was distributed as distinct small spots in the cytosol ( S6d and S6f Fig ) . These results indicated that EV71 alters the sub-cellular distribution of ASC . The effect of 3D on the sub-cellular distribution of NLRP3 was then determined in TPA-differentiated THP-1 macrophages infected with control lentivirus or 3D-encoding lentivirus . In the absence of 3D ( Fig 7Ga to 7Gd ) , NLRP3 was diffusely distributed in the cytoplasm ( Fig 7Ga and 7Gd ) . In the presence of 3D ( Fig 7Ge to 7Gh ) , NLRP3 was co-localized with 3D ( Fig 7Ge and 7Gf ) to form large spots in the cytoplasm ( Fig 7Gh ) . The association of 3D with the components of NLRP3 inflammasome was further examined in HEK293T cells co-transfected with plasmid expressing Flag-NLRP3 and plasmids expressing HA-3D , as indicated . In the absence of 3D , NLRP3 was diffusely distributed in the cytosol of HEK293T cells ( Fig 7Ha and 7Hd ) ; whereas in the presence of 3D , NLRP3 was co-localized with 3D ( Fig 7He and 7Hf ) to form large spots in the cytosol ( Fig 7He ) . The association of 3D with the components of NLRP3 inflammasome was also determined in RD cells co-transfected with plasmid expressing Flag-NLRP3 and plasmids expressing GFP-3D , as indicated . Similarly , in the absence of 3D , NLRP3 was diffusely distributed in the cytosol of RD cells ( Fig 7Ia and 7Id ) ; but in the presence of 3D , NLRP3 was co-localized with 3D ( Fig 7Ie and 7If ) to form large spots in the cytosol ( Fig 7Ie ) . These results revealed that 3D regulates the sub-cellular distribution of NLRP3 through interacting with NLRP3 in TPA-differentiated THP-1 macrophages , HEK293T cells , and RD cells . Taken together , we demonstrated that EV71 3D protein is associated with the major components ( NLRP3 , ASC , and pro-Casp-1 ) of NLRP3 inflammasome through interacting with NLRP3 , and also revealed that 3D regulates the sub-cellular distributions of NLRP3 inflammasome by binding to NLRP3 and forming a 3D-NLRP3 complex . Thus , these results suggested that 3D may regulate the assembly of NLRP3 inflammasome . In the absence of stimulation , NLRP3 is homo-oligomerized to form inactive preassembled complexes , which undergo conformational changes to form active inflammasome complexes containing ASC upon stimulation [31] . Since 3D protein is associated with the components of inflammasome through interacting with NLRP3 , we speculated that 3D may facilitate the assembly of inflammasome by direct binding to NLRP3 after EV71 infection . To verify this hypothesis , HEK293T cells were co-transfected with plasmids expressing Flag-NLRP3 , pcdna3 . 1 ( + ) -ASC , and HA-3D , as indicated . ASC was recruited by NLRP3 ( Fig 8A , lane 2 ) , and NLRP3-mediated recruitment of ASC was significantly enhanced by 3D ( Fig 8A , lane 3 ) . The oligomerization of ASC is critical for Casp-1 activation and inflammasome function [24] . Therefore , we further investigated the effect of 3D on the oligomerization of ASC . HEK293T cells were co-transfected with plasmids expressing pcdna3 . 1 ( + ) -ASC , Flag-NLRP3 , and HA-3D , as indicated . The results revealed that oligomerization of ASC was detected in the presence of ASC ( Fig 8B , lane 2 ) , enhanced by NLRP3 ( Fig 8B , lane 3 ) , and not affected by 3D in the absence of NLRP3 ( Fig 8B , lane 4 ) . However , in the presence of NLRP3 , oligomerization of ASC was significantly stimulated by 3D ( Fig 8B , lane 5 ) . Therefore , these results suggested that 3D facilitates the assembly of NLRP3 inflammasome through interacting with NLRP3 . It has been demonstrated that the localization of NLRP3 as spots is a character of the formation of inflammasome complex [6] . Thus , the effect of EV71 3D on the formation of NLRP3 inflammasome complex was explored in HEK293T cells co-transfected with plasmids encoding GFP , GFP-3D , Flag-NLRP3 , and/or HA-ASC , as indicated . In the absence of 3D and ASC ( Fig 8Ca to 8Ce ) , NLRP3 was diffusely distributed in the cell cytosol ( Fig 8Cb and 8Ce ) . In the presence of 3D and absence of ASC ( Fig 8Cf to 8Cj ) , NLRP3 was co-localized with 3D ( Fig 8Cf and 8Cg ) and distributed as large spots in the cell cytosol ( Fig 8Cj ) . In the absence of 3D and presence of ASC ( Fig 8Ck to 8Co ) , NLRP3 was co-localized with ASC ( Fig 8Ck and 8Ci ) and distributed as large spots in the cell cytosol ( Fig 8Co ) . Interestingly , in the presence of both 3D and ASC ( Fig 8Cp to 8Ct ) , NLRP3 was co-localized with 3D ( Fig 8Cp and 8Cq ) and ASC ( Fig 8Cq and 8Cr ) and distributed as large spots in the cytosol of transfected cells ( Fig 8Ct ) . These results clearly showed that 3D , NLRP3 , and ASC together formed a ring-like structure in the cytosol ( Fig 8Cp , 8Cq , 8Cr and 8Ct ) . More interestingly , we observed that in the 3D-NLRP3-ASC complex , 3D ( as indicated in green ) was located inside in the ring-like structure , followed by the 3D-NLRP3 complex ( green + red , as indicated in yellow ) and the NLRP3-ASC complex ( red + cyan , as indicated in white ) , and finally ASC ( as indicated in cyan ) was located outside the ring-like structure ( Fig 8Cu , enlarged ) . Moreover , the formation of 3D-NLRP3-ASC complex was also observed in HEK293T cells co-transfected with plasmids expressing GFP-3D , Flag-NLRP3 , and HA-ASC . The results confirmed that 3D , NLRP3 , and ASC proteins together were indeed distributed as large spots in the cytosol and formed a ring-like structure ( Fig 8De ) . In this ring-like structure , 3D was located inside ( Fig 8Da ) , followed by NLRP3 in the middle ( Fig 8Db ) , and ASC was located outside ( Fig 8Dc ) . These results suggested that 3D facilitates the assembly of NLRP3 inflammasome , since it has been demonstrated that the inflammasome complex is assembled by the formation of a ring-like organization [32] . Taken together , we demonstrated that 3D facilitates the assembly of NLRP3 inflammasome complex by the formation of a “3D-NLRP3-ASC” structure through direct binding to NLRP3 .
The assembly of inflammasomes is critical for the host innate immune response against pathogen infection [6] . The best-characterized inflammasomes is the NLRP3 inflammasome , which activates the maturation of IL-1β through promoting the cleavage of pro-Casp-1 to generate active subunits , p20 and p10 [14] . NLRP3 inflammasome regulates innate immunity in responding to several RNA viruses , including influenza A virus ( IAV ) [33] , encephalomyocarditis virus ( EMCV ) and vesicular stomatitis virus ( VSV ) [34] , measles virus ( MV ) [35] , myxoma virus ( MYXV ) [36] , adenovirus ( Ad ) [37] , West Nile virus ( WNV ) [38] , Rabies virus ( RV ) [39] , Herpes simplex virus 1 ( HSV-1 ) [40] , Rift valley fever virus ( RVFV ) [41] , human T-lymphotropic virus 1 ( HTLV-1 ) [42] , and EV71 [16] . However , the mechanisms underlying the assembly of NLRP3 inflammasome regulated by viruses have not been reported . In this study , we revealed a novel mechanism by which EV71 facilitates the assembly of NLRP3 inflammasome ( Fig 9 ) . We showed that EV71 stimulates the production of Casp-1 and the release of IL-1β in macrophages and PBMCs . EV71 induces IL-1β secretion only in the presence of all components ( NLRP3 , pro-Casp-1 , and ASC ) of NLRP3 inflammasome , and knock-down of the components attenuates EV71-induced activation of IL-1βwhich demonstrated that EV71 activates IL-1β through regulating NLRP3 inflammasome complex . The activation of NLRP3 inflammasome is regulated by two pathways in response to pathogen infection [7] . The first pathway leads to the activation of nuclear factor-kappa B ( NF-κB ) that subsequently induces the production of pro-IL-1β , NLRP3 , and ASC . The second pathway is initiated by the assembly of inflammasome , resulting in the activation of Casp-1 and the maturation of IL-1β . IL-1β acts as an important mediator of inflammation to stimulate the recruitment and activation of immune cells and the production of many secondary pro-inflammatory cytokines [15] . We further demonstrated that UV-inactivated EV71 and heat-inactivated EV71 fail to activate IL-1β and Casp-1 , suggesting that the replication of EV71 is necessary for the activation of IL-1β . Unlike HCV genomic RNA and influenza A virus RNA , EV71 genomic RNA is unable to activate NLRP3 inflammasome . In addition , the protein synthesis inhibitor CHX blocks EV71-mediated release of IL-1βindicating that viral protein synthesis is also required for the activation of IL-1β . Previous study had demonstrated that several viral proteins inhibit NLRP3 inflammsome-mediated IL-1β secretion , including MV V protein [35] , IAV NS1 and M2 proteins [43 , 44] , and encephalomyocarditis virus ( EMCV ) 2B protein [28] . Our detailed studies revealed that EV71 nonstructural proteins 3D are involved in the regulation of NLRP3 inflammasome . We demonstrated that EV71 3D plays a stimulatory role in the regulation of NLRP3 inflammasome . The 3D protein acts as a viral RdRp and plays an essential role in viral negative-strand synthesis and VPg uridylylation [45] . The VPg and uridylylated forms of VPg ( VPg-pUpU ) prime the initiation of RNA replication [46] . Many investigators have tried to inhibit the activity of 3D protein and thereby inhibit viral replication [30 , 47 , 48] . It is also involved in the regulation of cell S-phase arrest [49] and immune response [50] . However , the role of 3D in the regulation of NLRP3 inflammasome has not been reported . Thus , 3D function in the activation of NLRP3 inflammasome and the mechanism underlying such regulation were intensively investigated in this study . Interestingly , we showed that EV71 2A and 3C repress the production and secretion of IL-1β through attenuating NLRP3 inflammasome , which is consistent with a previous report [16] . More interestingly , we demonstrated that EV71 3D is the only viral protein that stimulates the secretion of IL-1β . We further revealed that 3D can directly bind with NLRP3 . We revealed that 3D interacts with NLRP3 LRR domain based on the result of yeast two-hybrid analysis , and further verified the interaction of 3D with NLRP3 by several approaches , including CoIP , GST pull down , and immunofluorescence assays . NLRP3 contains three important domains , PYD , NACHT , and LRR . We confirmed that 3D can interact with NACHT and LRR domains of NLRP3 . Upon activation , NLRP3 is oligomerized that leads to the interaction of NLRP3 with ASC PYRIN domain , which subsequently recruits pro-Casp-1 through CARD domain to permit the auto-cleavage and formation of active Casp-1 ( p10/p20 ) [7] . Although 3D was not interacted directly with ASC and pro-Casp-1 , it is associated with NLRP3 , ASC , and pro-Casp-1 , through interacting with NLRP3 . To our knowledge , there was no report describing a direct interaction between 3D and NLRP3 . Thus , we at the first time demonstrated that a viral RdRp interacts directly with NLRP3 to activate the inflammasome . Moreover , 3D alters the sub-cellular distributions of NLRP3 , ASC , and pro-Casp-1 , and is co-localized with the inflammasome components to form large spots through binding to NLRP3 . It has been demonstrated that distribution of NLRP3 as spots is a character of the formation of inflammasome complex [6] . Thus , our results suggested that 3D may regulate the assembly of NLRP3 inflammasome . In the absence of stimulation , NLRP3 is homo-oligomerized , which undergo conformational changes to form active inflammasome complex by interacting ASC upon stimulation [31] . 3D up-regulates the association of NLRP3 with ASC , suggesting that it is involved in the formation of inflammasome complex . The oligomerization of ASC is critical for Casp-1 activation and inflammasome function [24] . In the presence of NLRP3 , oligomerization of ASC is significantly stimulated by 3D , indicating that 3D may facilitate the assembly of NLRP3 inflammasome through interacting with NLRP3 . More interestingly , 3D , NLRP3 , and ASC together formed a ring-like structure , in which 3D interacts with NLRP3 that subsequently interacts with ASC . We observed that in the ring-like complex , 3D is located inside , followed by the 3D-NLRP3 complex and then the NLRP3-ASC complex , and finally ASC is located outside of the structure . It has been reported that the inflammasome complex is assembled by the formation of a ring-like organization [32] . Taken together , we demonstrated that 3D facilitates the assembly of NLRP3 inflammasome complex by the formation of a “3D-NLRP3-ASC” structure through direct binding to NLRP3 . In conclusion , we revealed a novel mechanism by which EV71 stimulates the activation of NLRP3 inflammasome by the virus-encoded 3D RNA polymerase . More importantly , 3D interacts directly with NLRP3 to facilitate the assembly of NLRP3 inflammasome complex by forming a “3D-NLRP3-ASC” ring-like structure . During the formation of the special structure , 3D binds to the LRR domain of NLRP3 that subsequently interacts with ASC through the PYRIN domain , ASC in turn binds to pro-Casp-1 by the CARD domain and activates Casp-1 ( p20/p10 ) , which finally stimulates the cleavage and release of IL-1β ( p17 ) . Thus , in this specific ring-like structure , the viral protein sites in the center , followed by NLRP3 in the middle and ASC then locates outside . IL-1β acts as an important mediator of inflammation by stimulating the activation of immune cells and the production of many secondary pro-inflammatory cytokines [15] , which play important roles in the development of inflammation and associated diseases [51] . Thus , this study discovers a new role of 3D as an important regulator in the activation of inflammatory response , reveals a novel mechanism underlying the regulation of inflammasome assembly mediated by viral invasion , and would provide new insights into development of agent for the treatment and prevention of viral associated inflammation and diseases .
Blood samples of healthy donors were randomly collected from Wuhan blood donation center ( Wuhan , China ) . To isolate peripheral blood mononuclear cells ( PBMCs ) , blood cells were separated from blood samples and diluted in RPMI-1640 purchased from Gibco ( Grand Island , NY , USA ) . Diluted blood cells ( 5 ml ) were added gently to a 15 ml centrifuge tube with 5 ml lymphocyte separation medium ( #50494 ) purchased from MP Biomedicals ( California , USA ) , and centrifuged at 2 , 000×g for 10 min at room temperature ( RT ) . The middle layer was transferred to another new centrifuge tube and diluted with RPMI-1640 . The remaining red blood cells were removed using red blood cell lyses buffer purchased from Sigma-Aldrich ( St . Louis , MO , USA ) . The pure PBMCs were centrifuged at 1 , 500×g for 10 min at RT and cultured in RPMI-1640 . The study was conducted according to the principles of the Declaration of Helsinki and approved by the Institutional Review Board of the College of Life Sciences , Wuhan University in accordance with its guidelines for the protection of human subjects . The Institutional Review Board of the College of Life Sciences , Wuhan University , approved the collection of blood samples for this research , in accordance with guidelines for the protection of human subjects . Written informed consent was obtained from each participant . Human rhabdomyosacroma cell line ( RD ) and human embryonic kidney cell line ( HEK 293T ) were purchased from American Type Culture Collection ( ATCC ) ( Manassas , VA , USA ) . Human monocytic cells ( THP-1 ) was a gift from Dr . Bing Sun of Institute of Biochemistry and Cell Biology , Shanghai Institute for Biological Sciences . THP-1 cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) , 100 U/ml penicillin , and 100 μg/ml streptomycin sulfate . RD and HEK293T cell lines were cultured in Dulbecco modified Eagle medium ( DMEM ) purchased from Gibco ( Grand Island , NY , USA ) supplemented with 10% fetal bovine serum ( FBS ) , 100 U/ml penicillin , and 100 μg/ml streptomycin sulfate . Cells were maintained in an incubator at 37°C in a humidified atmosphere of 5% CO2 . Lipopolysaccharide ( LPS ) , ATP , phorbol-12-myristate-13-acetate ( TPA ) , and dansylsarcosine piperidinium salt ( DSS ) were purchased from Sigma-Aldrich ( St . Louis , MO , USA ) . RPMI 1640 and Dulbecco modified Eagle medium ( DMEM ) were obtained from Gibco ( Grand Island , NY , USA ) . Nigericin was obtained from InvivoGene Biotech Co . , Ltd . ( San Diego , CA , USA ) . Antibody against Flag ( F3165 ) and monoclonal mouse anti-GAPDH ( G9295 ) were purchased from Sigma . Monoclonal rabbit anti-NLRP3 ( D2P5E ) , monoclonal rabbit anti-IL-1β ( D3U3E ) , monoclonal rabbit anti-caspase-1 ( catalog no . 2225 ) and monoclonal rabbit anti-eIF4G ( C45A4 ) were purchased from Cell Signaling Technology ( Beverly , MA , USA ) . Monoclonal mouse anti-ASC ( sc-271054 ) and polyclonal rabbit anti-IL-1β ( sc-7884 ) were purchased from Santa Cruz Biotechnology ( Santa Cruz , CA , USA ) . Monoclonal mouse anti-NLRP3 ( ALX-804-818 ) was purchased from Enzo Life Sciences ( Shang Hai , China ) to detection endogenous NLRP3 in THP1 cells by immunofluorescence microscopy . Polyclonal rabbit anti-3D antibody was produced by ABclonal Technology ( Wuhan , China ) . Lipofectamine 2000 , normal rabbit IgG and normal mouse IgG were purchased from Invitrogen Corporation ( CA , USA ) . In this study we used the Xiangyang strain of EV71 ( GenBank accession number JN230523 . 1 ) , which was previously isolated by our group [52 , 53] . The virus was adsorbed at 37°C for 2 h and the unbound virus was washed away . Infected cells were cultured in fresh medium supplemented with 2% FBS . For the preparation of UV-inactivated EV71 , the virus was dispersed in a tissue culture dish , and a compact UV lamp was placed directly above the dish for 30 min . For the preparation of heat-inactivated EV71 , the virus was incubated at 65°C for 30 min to completely inhibit the activity of EV71 . Virus titration was performed using RD cells in 96-well plates and expressed as the 50% tissue culture infectious dose ( TCID50 ) per unit volume . Sendai virus ( SeV ) strain was a gift from Dr . Hongbing Shu of Wuhan University . HCV genotype 2a strain JFH-1 was kindly provided by Takaji Wakita . RD cells were infected with EV71 virus at an infection of 0 . 1 PFU/cell . The supernatant was collected when cells showed maximal cytopathic effect from viral infection , centrifuged at 2 , 500 rpm for 30min , and then passed through 0 . 4μm filters . We used the E . Z . N . Z . viral RNA Kit for the isolation of EV71 virus RNA from the cell culture supernatant . The concentration of viral RNA was measured by NanoDrop 2000 which was purchased from Thermo scientific . THP-1 cells were differentiated to macrophages with 60 nM phorbol-12-myristate-13-acetate ( TPA ) for 12–14 h , and cells were cultured for 24 h without TPA . The differentiated cells were then stimulated in 6 cm plates with EV71 virus , Lipopolysaccharide ( LPS ) , Nigericin , or ATP . Supernatants were collected for measurement of IL-1β by ELISA . Cells were harvested for real-time PCR or immunoblot analysis . The cDNAs encoding human NLRP3 , ASC , pro-Casp-1 , and IL-1β were obtained by reverse transcription of total RNA from TPA-differentiated THP-1 cells , followed by PCR using specific primers . The cDNAs were sub-cloned into pcDNA3 . 1 ( + ) and pcDNA3 . 1 ( + ) -3×Flag vector . The pcDNA3 . 1 ( + ) -3×Flag vector was constructed from pcDNA3 . 1 ( + ) vector through inserting the 3×Flag sequence between the NheI and HindIII site . The primers used in this study are shown in S1 Table . To construct plasmids expressing EV71 proteins 2A , 2B , 3C , 3A , 3C , and 3D , corresponding fragments of EV71 cDNA were cloned into pEGFPC1 between the HindIII and SalI sites , resulting in green fluorescent protein ( GFP ) fusion protein . To construct pCAggs-HA-3D , the EV71 3D region was sub-cloned into pCAggs-HA vector using the EcoRI and KpnI sites . To construct pGEX6p-1-3D , the EV71 3D region was sub-cloned into pGEX6p-1 vector using BamHI and EcoRI sites . The PYRIN , NACHT , and LRR domain of NLRP3 protein was cloned into pcDNA3 . 1 ( + ) -3×Flag vector using specific primers shown in S1 Table . The targeting sequences of shRNAs for the human NLRP3 , ASC , and caspase-1 were as follows: sh-NLRP3: 5’-CAGGTTTGACTATCTGTTCT-3’; sh-ASC: 5’-GATGCGGAAGCTCTTCAGTTTCA-3’; sh-caspase-1: 5’-GTGAAGAGATCCTTCTGTA-3’ . A PLKO . 1 vector encoding shRNA for a negative control ( Sigma-Aldrich , St . Louis , MO , USA ) or a specific target molecule ( Sigma-Aldrich ) was transfected into HEK293T cells together with psPAX2 and pMD2 . G with Lipofectamine 2000 . We using the 3*Flag sequence to replace the GFP protein in the pLenti CMV GFP Puro vector ( Addgene , 658–5 ) for adding some Restriction Enzyme cutting site ( XbaI-EcoRV-BstBI-BamHI ) before the 3×Flag tag . Then the pLenti vector encoding EV71 3D protein was transfected into HEK293T cells together with psPAX2 and pMD2 . G with Lipofectamine 2000 . The primers were shown in S1 Table . Culture supernatants were harvested 36 h and 60 h after transfection and then centrifuged at 2 , 200rpm for 15 min . THP-1 cells were infected with the supernatants contain lentiviral particles in the presence of 4 μg/ml polybrene ( Sigma ) . After 48 h of culture , cells were selected by 1 . 5 μg/ml puromycin ( Sigma ) . The results of each sh-RNA-targeted protein and the lenti-3D protein were detected by real-time PCR and/or immunoblot analysis . The concentrations of IL-1β in culture supernatants were measured by ELISA kit ( BD Biosciences , San Jose , CA ) . The supernatant of the cultured cells was collected for 1 ml in the cryogenic vials ( Corning ) . The supernatant was frozen in -80°C for 4 h . The Rotational Vacuum concentrator machine that was purchase from Martin Christ was used for the freeze drying . The drying product was dissolved in 100 μl PBS and mixed with SDS loading buffer for western blotting analysis with antibodies for detection of activated caspase-1 ( D5782 1:500; Cell Signaling ) or mature IL-1β ( Asp116 1:500; Cell Signaling ) . Adherent cells in each well were lysed with the lysis buffer described below , followed by immunoblot analysis to determine the cellular content of various protein . The HEK293T whole-cell lysates were prepared by lysing cells with buffer ( 50 mM Tris-HCl , pH7 . 5 , 300 mM Nacl , 1% Triton-X , 5 mM EDTA , and 10% glycerol ) . The TPA-differentiated THP-1 cells lysates were prepared by lysing cells with buffer ( 50 mM Tris-HCl , pH7 . 5 , 150 mM Nacl , 0 . 1% Nonidetp40 , 5 mM EDTA , and 10% glycerol ) . Protein concentration was determined by Bradford assay ( Bio-Rad , Hercules , CA , USA ) . Cultured cell lysates ( 30 μg ) were electrophoresed in an 8–12% SDS-PAGE gel and transferred to a PVDF membrane ( Millipore , MA , US ) . PVDF membranes were blocked with 5% skim milk in phosphate buffered saline with 0 . 1% Tween 20 ( PBST ) before being incubated with the antibody . Protein band were detected using a Luminescent image Analyzer ( Fujifilm LAS-4000 ) . The HEK293T whole-cell lysates were prepared by lysing cells with buffer ( 50 mM Tris-HCl , pH7 . 5 , 300 mM Nacl , 1% Triton-X , 5 mM EDTA , and 10% glycerol ) . The TPA-differentiated THP-1 cells lysates were prepared by lysing cells with buffer ( 50 mM Tris-HCl , pH7 . 5 , 150 mM Nacl , 0 . 1% Nonidetp40 , 5 mM EDTA , and 10% glycerol ) . Lysates were immunoprecipitated with control mouse immunoglobulin G ( IgG ) ( Invitrogen ) or anti-Flag antibody ( Sigma , F3165 ) with Protein-G Sepharose ( GE Healthcare , Milwaukee , WI , USA ) . TPA-differentiated THP-1 cells was cultured or infected by the EV71 ( MOI = 20 ) virus for 24 h . Cells were fixed in 4% paraformaldehyde at room temperature for 15 min . After being washed three times with PBS , permeabilized with PBS containing 0 . 1% Triton X-100 for 5 min , washed three times with PBS , and finally blocked with PBS containing 5% BSA for 1 h . The cells were then incubated with the monoclonal mouse IgG1 anti-NLRP3 antibody ( ALX-804-818-C100; Enzo life Sciences ) and the polyclonal rabbit anti-3D antibody ( ABclonal technology ) overnight at 4°C , followed by incubation with FITC-conjugate donkey anti-mouse IgG and Dylight 649-conjugate donkey anti-rabbit IgG ( Abbkine ) for 1 h . After washing three times , cells were incubated with DAPI solution for 5 min , and then washed three more times with PBS . Finally , the cells were analyzed using a confocal laser scanning microscope ( Fluo View FV1000; Olympus , Tokyo , Japan ) . Total RNA was extracted with TRIzol reagent ( Invitrogen ) , following the manufacturer’s instructions . Real-time quantitative RT-PCR was performed using the Roche LC480 and SYBR RT-PCR kits ( DBI Bio-science , Ludwigshafen , Germany ) in a reaction mixture of 20 μl SYBR Green PCR master mix , 1 μl DNA diluted template , and RNase-free water to complete the 20 μl volume . Real-time PCR primers were designed by Primer Premier 5 . 0 and their sequences were was as follows: VP1 forward , 5’-CCCTTTAGTGGTTAGGATTT-3’ , VP1 reverse , 5’-CACCAGTTGGTTTAATGGAG-3’; NLRP3 forward , 5’-AAGGGCCATGGACTATTTCC-3’ , NLRP3 reverse , 5’-GACTCCACCCGATGACAGTT-3’; ASC forward , 5’-AACCCAAGCAAGATGCGGAAG-3’ , ASC reverse , 5’-TTAGGGCCTGGAGGAGCAAG-3’; caspase-1 forward , 5’-TCCAATAATGCAAGTCAAGCC-3’ , caspase-1 reverse , 5’-GCTGTACCCCAGATTTTGTAGCA-3’; IL-1β forward , 5’-CACGATGCACCTGTACGATCA-3’ , IL-1β reverse , 5’-AGCACTGAAAGCATGA-3’ , TNF-α forward , 5’-GGGTTTGCTACAACATGG-3’ , TNF-α reverse , 5’-AAGAACTTAGATGTCAGTGC-3’ , IL-8 forward , 5’-ACTTCTCCACAACCCTCTGC-3’ , IL-8 reverse , 5’-GTTGCTCCATATCCTGTCCCT-3’; GAPDH forward , 5’-AAGGCTGTGGGCAAGG-3’ , GAPDH reverse , 5’-TGGAGGAGTGGGTGTCG-3’ . The construct pGEX6p-1-3D plasmid and pGEX6p-1-LRR were transfected into Escherichia coli strain BL21 . After growing in LB medium at 3°C until the OD600 reached 0 . 6–0 . 8 , Isopropyl β-D-1-thiogalactopyranoside ( IPTG ) was added to a final concentration of 0 . 1 mM and the cultures grew for an additional 16 h at 16°C for GST-3D protein . Isopropyl β-D-1-thiogalactopyranoside ( IPTG ) was added to a final concentration of 1 mM and the cultures grew for an additional 4 h at 37°C for GST-LRR protein . And then the GST protein , GST-3D protein and GST-LRR protein were purified from the E . coli bacteria . For GST-3D pull-down assay , glutathione-Sepharose beads ( Novagen ) were incubated with GST-3D or GST protein . After washed with phosphate-buffered saline ( PBS ) , these beads were incubated with cell lysates from HEK293T which were transfected with plasmids encoding Flag-NLRP3 , Flag-ASC and Flag-pro-caspase-1 for 4 h at 4°C . The precipitates were washed three times , boiled in 2×SDS loading buffer , separated by 10% SDS-PAGE , immunoblotted with anti-GST , anti-Flag , and ant-ASC antibody . It was the same for the GST-LRR pull down assay . Saccharomyces cerevisiae strain AH109 , control vectors pGADT7 , pGBKT7 , pGADT7-T , pGBKT7-lam , and pGBKT7-p53 were purchased from Clontech ( Mountain View , CA , USA ) . Yeast strain AH109 was co-transformed with the combination of the pGADT7 and the pGBKT7 plasmids . Transformed yeast cells containing both plasmids were first grown on SD-minus Trp/Leu plates ( DDO ) to maintain the two plasmids and then were sub-cloned replica plated on SD-minus Trp/Leu/Ade/His plate ( QDO ) . The TPA-differentiated THP-1 cells were lysed by buffer ( 50 mM Tris , pH7 . 5 , 150 mM Nacl , 1% Nonidetp40 , 5 mM EDTA , and 10% glycerol ) at 4°C . The transfected HEK293T cells were lysed by buffer ( 50 mM Tris-HCl , pH7 . 5 , 300 mM Nacl , 1% Triton-X , 5 mM EDTA , and 10% glycerol ) . Lysates were centrifugated at 6000rpm for 15 min . The supernatants of the lysates were mixed with SDS loading buffer for western blot analysis with antibody against ASC . The pellets of the lysates were washed with PBS for three times and cross-linked using fresh DSS ( 2 mM , sigma ) at 37°C for 30 min . The cross-linked pellets were then spanned down and mixed with SDS loading buffer for western blotting analysis . All experiments were reproducible and repeated at least three times with similar results . Parallel samples were analyzed for normal distribution using Kolmogorov-Smirnov tests . Abnormal values were eliminated using a follow-up Grubbs test . Levene’s test for equality of variances was performed , which provided information for Student’s t-tests to distinguish the equality of means . Means were illustrated using histograms with error bars representing the SD; a P value of <0 . 05 was considered statistically significant . | The immune system protects the infected host and clears the invading pathogens . An important part of the innate immune response is the activation of NLRP3 inflammasome , which is induced upon exposure to pathogens . Activated inflammasome subsequently regulates the maturation of IL-1β that plays an important role in inflammatory response . Enterovirus 71 ( EV71 ) is a highly contagious virus causing hand-foot-mouth disease ( HFMD ) , meningoencephalitis , neonatal sepsis , and even fatal encephalitis in children by inducing many pro-inflammatory cytokines . Although NLRP3 inflammasome plays important role in regulating host immunity and viral infection , the assembly of NLRP3 inflammasome induced by viral infection is not known . In this study , we demonstrate that EV71 3D RNA polymerase activates NLRP3 inflammasome by binding to NLRP3 . More importantly , 3D was found to interact with NLRP3 to facilitate the assembly of inflammasome complex by forming a specific ring-like structure . Therefore , these findings demonstrate a new role of viral 3D polymerase in the activation of inflammatory response , and identify a novel mechanism underlying the regulation of inflammasome assembly in responding to pathogen infection , which would provide insights into the prevention and treatment of viral infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"enzyme-linked",
"immunoassays",
"molecular",
"probe",
"techniques",
"immunology",
"immunoblotting",
"physiological",
"processes",
"inflammasomes",
"molecular",
"biology",
"techniques",
"immunolog... | 2017 | EV71 3D Protein Binds with NLRP3 and Enhances the Assembly of Inflammasome Complex |
Most previous genetic epidemiology studies within the field of osteoporosis have focused on the genetics of the complex trait areal bone mineral density ( aBMD ) , not being able to differentiate genetic determinants of cortical volumetric BMD ( vBMD ) , trabecular vBMD , and bone microstructural traits . The objective of this study was to separately identify genetic determinants of these bone traits as analysed by peripheral quantitative computed tomography ( pQCT ) . Separate GWA meta-analyses for cortical and trabecular vBMDs were performed . The cortical vBMD GWA meta-analysis ( n = 5 , 878 ) followed by replication ( n = 1 , 052 ) identified genetic variants in four separate loci reaching genome-wide significance ( RANKL , rs1021188 , p = 3 . 6×10−14; LOC285735 , rs271170 , p = 2 . 7×10−12; OPG , rs7839059 , p = 1 . 2×10−10; and ESR1/C6orf97 , rs6909279 , p = 1 . 1×10−9 ) . The trabecular vBMD GWA meta-analysis ( n = 2 , 500 ) followed by replication ( n = 1 , 022 ) identified one locus reaching genome-wide significance ( FMN2/GREM2 , rs9287237 , p = 1 . 9×10−9 ) . High-resolution pQCT analyses , giving information about bone microstructure , were available in a subset of the GOOD cohort ( n = 729 ) . rs1021188 was significantly associated with cortical porosity while rs9287237 was significantly associated with trabecular bone fraction . The genetic variant in the FMN2/GREM2 locus was associated with fracture risk in the MrOS Sweden cohort ( HR per extra T allele 0 . 75 , 95% confidence interval 0 . 60–0 . 93 ) and GREM2 expression in human osteoblasts . In conclusion , five genetic loci associated with trabecular or cortical vBMD were identified . Two of these ( FMN2/GREM2 and LOC285735 ) are novel bone-related loci , while the other three have previously been reported to be associated with aBMD . The genetic variants associated with cortical and trabecular bone parameters differed , underscoring the complexity of the genetics of bone parameters . We propose that a genetic variant in the RANKL locus influences cortical vBMD , at least partly , via effects on cortical porosity , and that a genetic variant in the FMN2/GREM2 locus influences GREM2 expression in osteoblasts and thereby trabecular number and thickness as well as fracture risk .
Meta-analyses of genome-wide association studies ( GWAS ) have identified a large number of loci associated with areal bone mineral density ( aBMD ) [1]–[6] . aBMD is a complex trait , obtained from a 2-dimensional projectional scan of the given bone with dual x-ray absorptiometry ( DXA ) . Skeletal sites which are measured in this way , such as the lumbar spine and hip , comprise a mixture of cortical bone ( compact bone comprising the outer shell ) , and trabecular bone ( a network of thin interconnecting plates within the marrow cavity of vertebrae and the end of long bones ) . The lumbar spine has a relatively high proportion of trabecular bone , whereas the hip has a higher proportion of cortical bone . DXA-measured aBMD depends not only on bone cross-sectional size but also on apparent volumetric bone mineral density which is largely determined by trabecular microstructure and cortical thickness [7] . Although aBMD is the gold standard for diagnosing osteoporosis , it fails to provide a detailed skeletal phenotype necessary to discern traits such as trabecular volumetric BMD ( vBMD ) , cortical vBMD and bone microstructural parameters . Previous studies using DXA have demonstrated that age is a major predictor of fracture risk independent of aBMD . Although this aBMD independent effect of age has been attributed to poor bone “quality” , the structural basis for this remains unclear [8] . Age-related changes in bone include microstructural deterioration , such as trabecular perforation , thinning , and loss of connectivity , as well as increased cortical porosity [8] , [9] . Quantitative computed tomography ( QCT ) analysis has the capacity to reveal unique information about these bone traits . Standard peripheral QCT ( pQCT ) with a resolution of 500 µm has the advantage of being able to separately analyse trabecular and cortical vBMDs . The correlation between trabecular and cortical vBMDs is low ( rs 0 . 11 in the young adult men of the GOOD cohort; [10] ) , supporting the notion that the determinants of these two bone parameters differ . Cortical vBMD but not trabecular vBMD reflects material density while trabecular vBMD mainly is influenced by trabecular number and thickness . In addition , the correlations of these vBMD parameters with femoral neck aBMD are low ( cortical vBMD , rs 0 . 04 ) or moderate ( trabecular vBMD rs 0 . 65 ) , suggesting that cortical and trabecular vBMDs are at least partly influenced by genetic determinants not possible to identify by a GWAS of aBMD [10] . The heritability for trabecular vBMD has been reported to be as high as 59% while the heritability for cortical vBMD was slightly lower ( 40% ) [11] . GWAS have revealed differences in genetic associations with lumbar and hip aBMD , providing some evidence that cortical and trabecular bone have distinct genetic influences [2] . We have in a previous smaller-scale GWAS meta-analysis ( n = 1 , 934 ) identified a genetic variant in the RANKL locus to be significantly associated with cortical vBMD [10] . The genetic determinants of trabecular vBMD have not yet been evaluated using GWAS . High resolution pQCT ( HRpQCT ) not only allows the separation of the trabecular and cortical bone compartments but also the assessment of bone microstructure . HRpQCT has an isotrophic voxel size of 82 µm and shows excellent correlation with ex vivo μCT imaging ( resolution 20 µm or better ) [8] , [12] , [13] . Importantly , HRpQCT analysis recently demonstrated that younger and older subjects with the same aBMD differed in cortical porosity , a key parameter not captured by DXA [8] . The genetic determinants of trabecular and cortical bone microstructure parameters as analysed by HRpQCT are unknown . The objective of the present study was to identify genetic determinants of vBMDs and bone microstructure parameters separately for the cortical and trabecular bone compartments as analyzed by pQCT and HRpQCT . As our assembled discovery cohort was larger for the pQCT measurements ( cortical vBMD n = 5 , 878 , trabecular vBMD n = 2 , 500 ) than for the HRpQCT measurements ( n = 729 ) , we aimed to first identify genome-wide significant genetic variants for cortical and trabecular vBMDs separately and then to evaluate the impact of the identified variants on trabecular and cortical bone microstructure parameters in the HRpQCT cohort .
Table 1 displays the anthropometrics and bone traits for the four cohorts ( ALSPAC discovery , GOOD baseline discovery , YFS discovery , and MrOS Sweden replication ) evaluated . The association between cortical vBMD and trabecular vBMD was rather modest ( Spearman's rank correlation coefficient [rho] GOOD baseline r = 0 . 11 [10]; GOOD five year follow-up r = −0 . 01 ) . Separate GWA meta-analyses for cortical and trabecular vBMD were performed including all three discovery cohorts for cortical vBMD while trabecular vBMD was available in the YFS and GOOD cohorts . Inverse variance weighted fixed-effect model meta-analysis of study-specific results was performed . In the cortical vBMD GWA meta-analysis of the ALSPAC , GOOD and YFS cohorts there was little systematic inflation of test statistics ( Overall λ = 1 . 012 ( 1 . 023 for ALSPAC; 1 . 013 for GOOD; 1 . 023 for YFS ) ) , but a marked deviation from the null distribution amongst the lowest observed p-values ( Figure 1A ) . We identified genetic variants in four separate loci reaching genome-wide significance ( Figure 1B ) . The greatest evidence for association between genetic variation and cortical vBMD was seen for rs1021188 ( 0 . 15 SD decrease per C allele; p = 1 . 4×10−12 ) on chromosome 13 , slightly upstream of the RANKL gene ( TNFSF11; Table 2 , Figure 2A , Table S1 ) . The second strongest genetic signal for cortical vBMD ( rs271170; 0 . 11 SD decrease per T allele , p = 2 . 9×10−11 ) is a novel bone-related locus , located on chromosome 6 , upstream of LOC285735 ( Table 2 , Figure 2B , Table S1 ) . The third strongest signal ( rs7839059 , 0 . 10 SD decrease per A allele , p = 4 . 1×10−9 ) was located on chromosome 8 , upstream of OPG ( TNFRSF11B; Table 2 , Figure 2C , Table S1 ) . The fourth genome-wide signal ( rs6909279 , 0 . 09 SD decrease per allele G , p = 1 . 0×10−8 ) was located on chromosome 6 , in C6orf97 upstream and close to estrogen receptor-α ( ESR1; Table 2 , Figure 2D , Table S1 ) . We selected our top four regions and carried out analyses conditional on the most associated SNPs in each region . When conditioning on the most significant SNP in the RANKL region ( rs1021188 ) an additional suggestive signal ( rs17638544 close to AKAP1 and upstream of RANKL , p = 4 . 2×10−5 ) appeared , but did not achieve genome-wide significance ( Table 2 , Figure 2E , Table S1 ) . Using similar conditional analysis , no additional SNPs with an independent signal appeared in the other three evaluated regions ( p<5×10−5 ) . The RANKL , OPG and ESR1 regions have earlier been reported to be associated with aBMD in large scale GWA meta-analyses [1]–[6] . To evaluate if the identified SNPs associated with cortical vBMD in these regions are independent from the previously reported aBMD related SNPs , conditional analyses were performed ( RANKL region , rs1021188 and rs17638544 were conditioned on the known aBMD hit rs9533090; OPG region , rs7839059 was conditioned on the known aBMD hit rs2062377; ESR1 region , rs6909279 was conditioned on known aBMD hits rs7751941 and rs4869742; [2] ) . The two cortical vBMD RANKL signals ( rs1021188 and rs17638544 ) were distinct from the previously reported aBMD signal ( rs9533090; [2] ) in this region , supported by the fact that ( i ) rs9533090 was not significantly associated with cortical vBMD ( Figure 2A ) , ( ii ) adjustment for rs9533090 did not influence the associations for rs1021188 or rs17638544 with cortical vBMD and the two cortical vBMD signals displayed a low r2 ( <0 . 04 ) with rs9533090 ( Table S2 ) . It is difficult to determine if the identified cortical vBMD signal in the OPG region is separate from the previous reported aBMD signal in this region ( rs2062377; [2] ) as this previous aBMD signal also was significantly associated with cortical vBMD ( Figure 2C ) , the r2 between the two SNPs was 0 . 39 , and adjustment for rs2062377 slightly but not completely attenuated the association for rs7839059 with cortical vBMD ( Table S2 ) . The identified cortical vBMD SNP in the ESR1 region ( rs6909279 ) is independent from one of the previous reported aBMD signals ( rs7751941 ) while the other reported independent aBMD SNP in this region ( rs4869742 [2] ) displayed a relatively high r2 with rs6909279 ( r2 = 0 . 60 ) ( Figure 2D ) . However , adjustment for rs4869742 only slightly attenuated the association for rs6909279 with cortical vBMD ( Table S2 ) . Sex-specific analyses demonstrated that all five cortical vBMD SNPs were significantly associated with cortical vBMD in both men and women with effects in the same direction . However , the magnitude of the effect sizes differed significantly according to sex for three of the identified hits ( Figure 5A; rs1021188 in RANKL region , men 0 . 21 SD and women 0 . 06 SD decrease per C allele , p = 4 . 3×10−5; rs6909279 in ESR1 region , men 0 . 12 SD and women 0 . 05 SD decrease per G allele , p = 3 . 0×10−3; rs7839059 in OPG region , men 0 . 12 SD and women 0 . 07 SD decrease per A allele , p = 3 . 0×10−2 ) . The sex difference may explain the heterogeneity observed in the unstratified analyses for rs1021188 ( both sexes have a het p>0 . 05 ) . Significant heterogeneity is still observed in the males for rs17638544 ( het p = 0 . 016 ) . The genetic variant associated with trabecular vBMD ( rs9287237 ) was significantly associated with trabecular vBMD in men ( Figure 5B , 0 . 22 SD increase per T allele ) and a similar non-significant tendency in the same direction was observed in females ( 0 . 12 SD increase per T allele ) . It should be emphasized that for trabecular vBMD , we had a low statistical power to detect sex differences in effect sizes as the number of females included was rather low ( n = 865 ) . At the five year follow-up of the GOOD cohort both pQCT analysis , giving information about cortical and trabecular vBMD , and HRpQCT analyses , giving information about trabecular bone microstructure and cortical porosity , were available in the tibia for 729 subjects with genotype data available ( Table 4 ) . To determine the impact of the identified genome-wide significant cortical and trabecular vBMD signals for bone microstructure parameters , their associations with HRpQCT parameters were evaluated in the GOOD cohort . Trabecular vBMD as analysed by pQCT was strongly ( r = 0 . 94 ) associated with trabecular bone fraction ( BV/TV ) as analysed by HRpQCT . The pQCT-derived cortical vBMD was moderately inversely correlated to cortical porosity as analysed by HRpQCT ( r = −0 . 21 ) . Although there appeared to be no overlap in the identity of the genome-wide significant SNPs between cortical and trabecular vBMD , it is still possible that there are genetic variants lower down the distribution of tests statistics which do not meet the stringent criteria for genome-wide significance , but nevertheless affect both traits pleiotropically . In order to investigate this possibility we ran a bivariate REML analysis using the GCTA software package in the GOOD cohort , having both cortical and trabecular vBMDs measurements available [14] . GCTA estimated the genetic correlation between trabecular and cortical BMD as rG = 0 . 0 ( SE = 0 . 39 ) suggesting an absence of common genetic variants affecting both traits and consistent with our results from the genome-wide association analysis . However , we note that there are relatively few individuals in this analysis and consequently the standard errors on this estimate are very wide . In order to be more definitive with respect to the possible existence of pleiotropy one would need to perform the analysis in a much larger sample of individuals to yield precise estimates of the genetic correlation between the two traits . All five genome-wide significant vBMD SNPs were nominally significantly associated ( p<0 . 05 ) with both femoral neck and lumbar spine aBMD as provided in the public data release from the discovery phase ( n≅32 , 000 ) of the recent aBMD analyses from the GEFOS consortium ( Table 3; http://www . gefos . org/ ? q=content/data-release ) [2] . The direction of the effect was the same when comparing vBMDs and aBMD for four of the SNPs while it was opposite to the one described for aBMD for the cortical vBMD SNP rs271170 . When evaluating the 64 genome-wide significant aBMD SNPs recently identified by the GEFOS consortium [2] it was found that 15 of these were also significantly associated ( p<0 . 05 ) with cortical vBMD and 15 were significantly associated with trabecular vBMD . Four of these SNPs were associated with both cortical and trabecular vBMDs ( Table S4 ) . In an attempt to assess the underlying functional mechanism of our identified loci we examined their potential role in regulating gene expression using expression quantitative trait locus ( eQTL ) data from resting ( i . e . untreated ) and induced ( i . e . dexamethasone , BMP-2 and PGE2 treated ) primary human osteoblasts [15] , [16] . Expression of genes in close proximity to the five genome-wide significant SNPs ( defined as located within the gene ±250 kb ) was tested for association ( Table S5 ) . We found that the trabecular vBMD-associated SNP ( rs9287237 ) was the strongest SNP significantly associated ( P = 2 . 3×10−4 ) with expression of the nearby GREM2 gene . No significant effects on gene expression were noted at the additional four loci ( Bonferroni adjusted P>0 . 05 corresponding to 0 . 05/88 = 5 . 7×10−4; Table S5 ) . Overall , 388 men had at least one validated incident fracture after an average follow-up of 5 . 4 years in the MrOS Sweden cohort ( Table S6 ) . The trabecular vBMD SNP rs9287237 , but none of the four cortical vBMD SNPs , was significantly associated with risk of all fractures ( HR per extra T allele 0 . 75 , 95% confidence interval ( CI ) 0 . 60–0 . 93 ) and hip fractures ( HR per extra T allele 0 . 59 , 95% CI 0 . 36–0 . 98; Table S7 ) . In two centers of MrOS Sweden , Gothenburg and Malmö , overall , 1445 men had a spine X-ray at baseline , which identified 225 men with prevalent vertebral fractures ( 15 . 6%; Table S6 ) . The trabecular vBMD SNP rs9287237 , but none of the four cortical vBMD SNPS , was significantly associated with prevalent X-ray verified vertebral fractures ( OR per extra T allele 0 . 68 , 95% CI 0 . 50–0 . 94; Table S7 ) .
Osteoporosis is a common highly heritable skeletal disease characterized by reduced aBMD and deteriorated bone microstructure , resulting in an increased risk of fragility fracture [17] . Most previous genetic epidemiology studies have focused on the genetics of the complex trait aBMD , not being able to separate genetic determinants of trabecular vBMD , cortical vBMD and bone microstructure . However , all these structural traits are determinants of aBMD [7] . We , herein , provide evidence that the genetic determinants of cortical and trabecular vBMDs differ . Separate GWAS meta-analyses of cortical and trabecular vBMDs demonstrated that genetic variants in the RANKL , LOC285735 , OPG and ESR1 loci were associated with cortical vBMD while a genetic variant in the FMN2/GREM2 locus was associated with trabecular vBMD . Follow-up analyses in a HRpQCT cohort providing bone microstructural traits revealed that a genetic variant in the RANKL locus is associated with not only cortical vBMD but also with cortical porosity . The FMN2/GREM2 locus , on the other hand , is robustly associated with not only trabecular vBMD but also trabecular number and thickness as well as GREM2 expression in human osteoblasts and fracture risk . Using the same pQCT cohorts as evaluated in the present study , we have recently identified a missense variant in the WNT16 gene to be associated with cortical bone thickness [18] . Subsequent functional studies revealed that Wnt16 inactivated mice had reduced cortical bone thickness , providing strong evidence that WNT16 is an important regulator of cortical bone thickness . In the present discovery GWAS-meta-analysis , we identified genome-wide significant cortical vBMD signals in four loci and a trabecular vBMD signal in one locus and all these signals were significantly replicated in the MrOS Sweden cohort . One of the identified cortical vBMD loci ( LOC285735 ) and the identified trabecular vBMD locus ( FMN2 ) are novel bone-related loci while the remaining three identified cortical vBMD loci have been previously reported to be associated with aBMD . [2] , [6] , [19] . Although the number of subjects included in the present discovery GWAS meta-analyses of vBMD ( cortical vBMD n = 5878; trabecular vBMD n = 2500 ) was substantially lower than the number of subjects included in a recent large-scale aBMD GWA meta-analysis ( discovery cohort n = 32 , 961; [2] ) , the identification of novel bone-related loci in the trabecular and cortical vBMD analyses is not surprising as the correlations between aBMD on the one side and cortical and trabecular vBMDs on the other side are low/modest . In addition , the genetic variants associated with cortical and trabecular bone parameters , respectively , differed , and were also distinct from the WNT16 locus recently identified to be associated with cortical bone thickness [18] , underscoring the complexity of the genetics of bone parameters . Conditional analysis revealed a secondary signal in the RANKL locus and both these cortical vBMD signals were independent of the previously reported aBMD signal ( rs9533090; [2] ) in this region , demonstrating that separate signals within the same region can have an impact on different bone traits ( = allelic heterogeneity ) . RANKL exerts its biological effects on bone by stimulating osteoclast differentiation following interactions with its receptor , RANK; how distinct genetic pathways might influence this functionality in different ways , so as to influence distinct phenotypic traits , is currently unclear . Alternatively , one of these signals may be in LD with a marker at a different gene responsible for mediating the genetic effect in question , or else represent a variant which although trans to a structural gene , affects transcription at other sites [20] . The cortical vBMD SNPs rs7839059 ( TNFRSF11B locus ) was also nominally ( p<0 . 05 ) significantly associated with trabecular vBMD , although with less pronounced effect size , suggesting that this SNP does not exclusively have an impact on cortical bone . The present report describing two independent RANKL signals and one OPG signal with an impact on cortical vBMD provides further evidence that the RANK/RANKL/OPG axis affects the skeleton at least in part by influencing volumetric apparent density of cortical bone . It is tempting to speculate that changes in cortical vBMD contribute to the recent observations that the RANKL inhibitor denosumab reduces fracture risk [10] , [21] , [22] . Consistent with this possibility , administration of denosumab has been found to increase femoral cortical vBMD in mice with a knock-in of humanized RANKL [23] . The second strongest genetic signal for cortical vBMD was located on chromosome 6 ( rs271170 ) , 93 . 4 kb upstream of LOC285735 . This is a novel bone-related signal and further targeted sequencing efforts and functional studies are required to characterize this signal . Several clinical and preclinical studies have clearly demonstrated that ESR1 is an important regulator of both female and male bone health [24]–[28] but the present study is first to provide genetic evidence that this receptor influences the volumetric apparent density of cortical bone . This finding is of importance as Khosla and co-workers recently proposed that the main physiological target for estrogen in bone is cortical and not trabecular bone [24] . A significant signal ( rs9287237 ) for trabecular vBMD was identified on chromosome 1 located in the intron region of the FMN2 gene . The combined effect size of this signal was substantial with an increase of 0 . 19 SD per T allele . FMN2 is a gene that is expressed in oocytes and is required for progression through metaphase of meiosis 1 but it is not previously reported to influence the skeleton [29] . However , a genetic variant within FMN2 has been associated with coronary heart disease [30] . The rs9287237 SNP is located slightly ( 55 . 7 kb ) downstream of GREM2 ( = PRDC ) , which is an extracellular antagonist of bone morphogenetic proteins ( BMPs ) and it inhibits osteoblastic differentiation [31] , [32] , making it an alternative plausible candidate gene underlying the rs9287237 association with trabecular vBMD . Importantly , eQTL analyses in human osteoblasts demonstrated that the trabecular vBMD-associated SNP ( rs9287237 ) was significantly associated with expression of the nearby GREM2 gene , indicating that GREM2 is a strong candidate for mediating the trabecular vBMD association at rs9287237 . However , further targeted sequencing efforts and functional studies are required to characterize this signal . There are known sex differences in bone traits in mice [33]–[36] . Similarly , some genome-wide linkage analyses in humans have reported sex-specific results . In a whole–genome linkage analysis stratified by sex , sex-specific QTLs were found in the Framingham sample [37] . Furthermore , in a meta-analysis that included data from nine whole-genome linkage scans for aBMD , several sex-specific QTLs were observed [38] . To our knowledge there is only one reported genome-wide significant aBMD signal , located on the X-chromosome ( Xp22 . 31 ) , which displays significant sex heterogeneity [2] . This signal was only significant in men and the same signal was also shown to be associated with male serum testosterone levels [39] . Sex-specific analyses in the present study revealed that all identified cortical vBMD signals were significantly associated with cortical vBMD in both men and women with effects in the same direction . Nevertheless , the magnitude of the effect sizes differed significantly according to sex for three of the identified hits . Importantly , the effect sizes of the RANKL and ESR1 signals were more than three ( 0 . 21 SD vs . 0 . 06 SD ) and two ( 0 . 12 vs . 0 . 05 SD ) times larger , respectively , in men than in women . The smaller effect within females observed for rs1021188 in the RANKL region is mainly driven by ALSPAC , and there may be other reasons ( such as younger age ) why this study showed a smaller effect . However , the consistent results between ALSPAC and the YFS provide some evidence against the differences being driven primarily by age . The relative strong ESR1 signal in men supports experimental and clinical studies , demonstrating that estrogens are crucial for male bone health [24] , [25] , [27] , [40] . We examined genetic effects across cohorts encompassing a relatively broad age range , including 15 year old participants from ALSPAC who were still attaining peak bone mass , to older men from MrOS Sweden starting to show age-related bone loss . Inclusion of an older cohort had the advantage of providing an opportunity to study relationships with fracture risk . However , this design may have reduced the power to detect genetic associations by introducing greater heterogeneity . In contrast to aBMD , vBMD has been reported to change relatively little from adolescence to mid-life suggesting that analyses combining cohorts of different ages might be more informative when based on this trait [41] . However , recent follow up studies based on the GOOD cohort revealed substantial changes in cortical vBMD in the late teens and early twenties , at least in males [42] . Hence the suggestion that certain genetic associations with cortical vBMD were weaker in ALSPAC compared with other cohorts may reflect attenuation of effect during the consolidation of cortical bone whilst attaining peak bone mass . Age-related changes in bone include microstructural deterioration , such as trabecular perforation , thinning , and loss of connectivity , as well as increased cortical porosity [8] , [9] . These bone microstructural parameters are believed to have an aBMD-independent influence on fracture risk and they can be analyzed by HRpQCT . The present study is the first to identify genetic loci associated with cortical and trabecular bone microstructural parameters as analyzed by HRpQCT . The SNP in the RANKL region with the strongest association with cortical vBMD was also significantly associated with cortical porosity but , as expected , in the reverse direction . This finding suggests that a genetic variant in the RANKL locus influences cortical vBMD , at least partly , via effects on cortical porosity . Importantly , this signal in the RANKL region was independent from the previously reported aBMD signal in the same region [2] . Analyses of trabecular bone microstructure demonstrated that the trabecular vBMD SNP rs9287237 in the FMN2/GREM2 locus was significantly associated with several trabecular but not cortical bone microstructure parameters . When evaluated in the five-year follow-up visit in the GOOD cohort , each T allele of this SNP resulted in a substantial increase in trabecular vBMD ( 0 . 32 SD ) , trabecular bone fraction ( BV/TV 0 . 29 SD ) , trabecular number ( 0 . 15 SD ) , and trabecular thickness ( 0 . 18 SD ) . Thus , a genetic variant in the FMN2/GREM2 locus influences trabecular vBMD via substantial effects on both trabecular number and thickness . Although , the present study is the first to report on genetic variants associated with microstructural bone-parameters , the analyses were candidate-based as a follow-up of our initial cortical and trabecular vBMD GWA meta-analyses . In order to identify novel genetic loci for bone microstructural parameters in a hypothesis-free manner , well-powered HRpQCT cohorts with genome-wide genotype data available need to be established . We believe that our study provides strong evidence that previous large-scale GWA meta-analyses of the complex bone trait aBMD did not have the capability to identify a number of important loci with an impact on aspects of micro-architecture which may have important effects on fracture risk but be poorly reflected by overall aBMD measurements . We , therefore , propose that future well-powered pQCT and HRpQCT GWA meta-analyses of these specific bone structural traits will add useful information and might result in the identification of novel osteoporosis drug targets and provide novel aBMD-independent genetic markers for the prediction of fracture risk . The implication of our results suggesting that cortical and trabecular bone compartments are under distinct genetic control is consistent with the fact that patients with idiopathic osteoporosis may present with a predominantly trabecular or cortical bone phenotype [43] . Although the lumbar spine and hip both comprise a mixture of bone types , the former has a relatively high proportion of trabecular bone , whereas the hip has a higher proportion of cortical bone . Hence , patients presenting with a disproportionate decrease in lumbar spine aBMD , which are well recognized , presumably have greater reductions in trabecular compared to cortical BMD [44] . Further studies are required to determine whether genetic variation in the FMN2/GREM2 locus helps to explain this type of presentation . The genetic variant in the FMN2/GREM2 locus was associated with fracture risk and prevalent X-ray verified vertebral fractures in the MrOS Sweden cohort . However , further large-scale studies are required to replicate the fracture findings of this SNP . Collectively our data demonstrate that each extra T allele of rs9287237 is associated with decreased expression of the BMP antagonist GREM2 in osteoblasts , increased trabecular vBMD and decreased fracture risk . As previous in vitro studies have demonstrated that GREM2 inhibits osteoblast differentiation , we propose that rs9287237 is involved in the regulation of GREM2 expression in osteoblasts , which in turn regulates osteoblast differentiation and thereby the amount of trabecular bone and fracture risk [31] , [32] . In conclusion , we identified five genetic loci associated with trabecular or cortical vBMD . Two of these ( FMN2/GREM2 and LOC285735 ) are novel bone-related loci while the other three have previously been reported to be associated with aBMD . The genetic variants reported to be associated with cortical and trabecular bone parameters differed and were also distinct from the WNT16 locus recently identified to be associated with cortical thickness [18] , underscoring the complexity of the genetics of bone parameters . Finally , we propose that a genetic variant in the RANKL locus influences cortical vBMD , at least partly , via effects on cortical porosity , and that a genetic variant in the FMN2/GREM2 locus influences trabecular vBMD and fracture risk via substantial effects on both trabecular thickness and number .
All study participants provided informed written consent . Approval by local institutional review boards was obtained in all studies . SNPs associated with vBMD at the genome-wide significance level as reported here were tested for association with resting or induced gene expression of neighbouring gene transcripts , in primary human osteoblasts derived from 113 ( 51 female and 62 male donors , respectively ) unrelated Swedish donors . Detailed cell culture and analysis methods have been described in detail [15] , [16] . Briefly , expression profiling of untreated , dexamethasone , BMP-2 and PGE2-treated cells each with up to three biological replicates was performed using the Illumina HumRef-8 BeadChips according to the protocol supplied by the manufacturer . Genotyping for genotype-expression association was performed using Illumina HapMap 550 k Duo chip . Individuals with low genotyping rate and SNPs showing significant deviation from Hardy-Weinberg equilibrium ( P<0 . 05 ) were excluded . Similarly low frequency ( MAF<0 . 05 ) SNPs and SNPs with high rates of missing data were excluded . Genotypes from samples that passed quality control ( N = 103 ) were imputed for all SNPs ( n = 478 , 805 ) oriented to the positive strand from phased ( autosomal ) chromosomes of the HapMap CEU Phase II panel ( release 22 , build 36 ) using MACH 1 . 0 ) [55] . A cut-off of R2<0 . 3 were used to remove poorly imputed markers . Association of imputed genotypes using estimated genotype probabilities with nearby expression traits ( defined as ±250 kb window flanking the gene ) were performed using a linear regression model implemented in the MACH2QTL software with sex and age as covariates . Participants were followed for 5 . 4 years on average after the baseline examination . The follow-up time was recorded from the date of the baseline visit to the date of the first fracture , the date of death , or end of the present follow-up interval . When a subject sustained a first fracture at different sites during the follow-up , the various fractures and the follow-up time for each respective first fracture type were included in the analyses . Complete follow-up was possible because central registers covering all Swedish citizens were used to identify the subjects and the time of death for all subjects who died during the study , and these analyses were performed after the time of fracture validation . At the time of fracture evaluation , the computerized X-ray archives in Malmö , Göteborg , and Uppsala were searched for new fractures occurring after the baseline visit , using the unique personal registration number , which all Swedish citizens have . All fractures reported by the study subject after the baseline visit were confirmed by physician review of radiology reports . Fractures reported by the study subject , but not possible to confirm by X-ray analyses report , were not included in this study . All validated fractures and hip fractures were evaluated . Information about the type of trauma associated with the incident fractures was not available . Fracture rates were expressed as the number of subjects with first fractures per 1000 person-years ( Table S6 ) . In Gothenburg and Malmö , 1445 men had an X-ray of the lateral thoracic and lumbar spine at baseline . All vertebral fractures were evaluated by an expert radiologist . If the vertebral body had a reduced height of 3 mm or more compared with the vertebra above , it was classified as a vertebral fracture [56] . The ALSPAC ( n = 3382 ) , YFS ( n = 1558 ) and GOOD ( n = 938 ) discovery cohorts contributed to the cortical vBMD genome-wide meta-analysis while the YFS and GOOD discovery cohorts contributed to the trabecular vBMD genome-wide meta-analysis . We analyzed only those imputed SNPs which had a minor allele frequency of >0 . 01 and an r2 imputation quality score of >0 . 3 in all 3 sets ( n = 2 , 401 , 124 ) . We carried out genome-wide association analyses for cortical and trabecular vBMDs using additive linear regression in Mach2QTL for ALSPAC , ProbABEL [57] for YFS and Mach2QTL on GRIMP [19] for the GOOD analyses . We included age , sex , height and weight ( ln ) as covariates . We carried out meta-analyses of the results from the three cohorts using the inverse variance method in METAL . Standardized betas and standard errors from each study were combined using a fixed effect model which weights the studies using the inverse variance and applying genomic control to individual studies and the combined results . Genome-wide significance was taken to be p<5×10−8 . We also repeated the analyses in each of the three discovery cohorts , conditional on these top SNPs , to identify any additional independent associations in the regions . We selected one SNP for replication in the MrOS Sweden cohort from each independent region that had a p<5×10−8 as well as a secondary SNP from the RANKL region which appeared to influence cortical vBMD . Additive linear regression analyses were carried out for the associations between these SNPs and cortical and trabecular vBMDs in SPSS Statistics 17 . 0 for MrOS Sweden , using age , sex , height and weight ( ln ) as covariates . The results of all four cohorts were combined using a fixed effects inverse-variance meta-analysis in Stata ( version 11 . 2 ) . The SNPs showing evidence for heterogeneity ( as assessed by a chi-squared test ) were also meta-analysed using the DerSimonian & Laird random effects method . Correlations between bone traits in the GOOD cohort were tested and presented as Spearman's rank correlation coefficients ( rho ) . The difference of the allelic association effects between males and females was tested using a two sample z-test . Cox proportional hazards models were used to study the associations between SNPs and incident fractures . Prevalent vertebral fractures were analyzed using binary logistic regression models . In order to estimate the genetic correlation between cortical and trabecular vBMD we ran a bivariate REML analysis using the GCTA software package in the GOOD cohort , having both cortical and trabecular vBMDs measurements available [14] . This method essentially decomposes the covariance between the two traits into a part explained by common genetic variation that is tagged by markers on the SNP chip and a residual part that is not . It is then possible to use these estimates to estimate a genetic correlation between the two . | Osteoporosis is a common highly heritable skeletal disease characterized by reduced bone mineral density ( BMD ) and deteriorated bone microstructure , resulting in an increased risk of fracture . Most previous genetic epidemiology studies have focused on the genetics of the complex trait BMD , not being able to separate genetic determinants of the trabecular and cortical bone compartments and bone microstructure . The trabecular and cortical BMDs can be analysed separately by computed tomography . Therefore , we performed separate genome-wide association studies for trabecular and cortical BMDs , demonstrating that the genetic determinants of cortical and trabecular BMDs differ . Genetic variants in the RANKL , LOC285735 , OPG , and ESR1 loci were associated with cortical BMD , while a genetic variant in the FMN2/GREM2 locus was associated with trabecular BMD . Two of these are novel bone-related loci . Follow-up analyses of bone microstructure demonstrated that a genetic variant in the RANKL locus is associated with cortical porosity and that the FMN2/GREM2 locus is associated with trabecular number and thickness . We propose that a genetic variant in the RANKL locus influences cortical BMD via effects on cortical porosity , and that a genetic variant in the FMN2/GREM2 locus influences trabecular BMD and fracture risk via effects on both trabecular number and thickness . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"endocrinology",
"diabetes",
"and",
"endocrinology",
"genetics",
"and",
"genomics",
"clinical",
"genetics"
] | 2013 | Genetic Determinants of Trabecular and Cortical Volumetric Bone Mineral Densities and Bone Microstructure |
Heterozygous germline mutations and deletions in PHOX2B , a key regulator of autonomic neuron development , predispose to neuroblastoma , a tumor of the peripheral sympathetic nervous system . To gain insight into the oncogenic mechanisms engaged by these changes , we used zebrafish models to study the functional consequences of aberrant PHOX2B expression in the cells of the developing sympathetic nervous system . Allelic deficiency , modeled by phox2b morpholino knockdown , led to a decrease in the terminal differentiation markers th and dbh in sympathetic ganglion cells . The same effect was seen on overexpression of two distinct neuroblastoma-associated frameshift mutations , 676delG and K155X - but not the R100L missense mutation - in the presence of endogenous Phox2b , pointing to their dominant-negative effects . We demonstrate that Phox2b is capable of regulating itself as well as ascl1 , and that phox2b deficiency uncouples this autoregulatory mechanism , leading to inhibition of sympathetic neuron differentiation . This effect on terminal differentiation is associated with an increased number of phox2b+ , ascl1+ , elavl3− cells that respond poorly to retinoic acid . These findings suggest that a reduced dosage of PHOX2B during development , through either a heterozygous deletion or dominant-negative mutation , imposes a block in the differentiation of sympathetic neuronal precursors , resulting in a cell population that is likely to be susceptible to secondary transforming events .
Neuroblastoma is an embryonal malignancy of the peripheral sympathetic nervous system ( PSNS ) that arises from the developing neural crest and manifests as neoplasms in sympathetic ganglia or adrenal medulla . The oncogenic events culminating in neuroblastoma are thought to occur very early in development , consistent with the status of this tumor as the most common cancer of infants [1] . A defining feature of neuroblastic tumors is their broad spectrum of cellular differentiation , ranging from undifferentiated cells that indicate a poor prognosis to those showing greater differentiation and predicting a generally favorable outcome [2] . This heterogeneity suggests that dysregulated differentiation of sympathetic progenitor cells plays a key role in neuroblastoma pathogenesis . Direct evidence for this model comes from the identification of heterozygous germline mutations in the homeodomain transcription factor PHOX2B , a regulator of sympathetic neuronal differentiation [3]–[8] . Such mutations are also found in patients with other neural crest-derived disorders , including congenital central hypoventilation syndrome ( CCHS ) and Hirschprung's disease , characterized by absent or abnormal development of the noradrenergic neurons in the brain stem and colon , respectively [3] , [5] , [7] , [8] . The vast majority of embryonal tumors like neuroblastoma arise from aberrant genetic and epigenetic changes that control the survival , proliferation and differentiation of specific tissues . Hence , one way to decipher the tumorigenic contribution of germline changes in highly conserved genes such as PHOX2B is to understand how they perturb normal development . The sympathetic nervous system is derived from the neural crest , a multipotent embryonic structure consisting of a transient population of cells that migrate from the neural tube to the region of the dorsal aorta during development [9]–[11] . At the dorsal aorta , prespecified SOX10-positive sympathetic progenitors , under the influence of bone morphogenetic proteins ( BMPs ) , start the process of differentiation into noradrenergic neurons [12] . Phox2b and the basic helix-loop-helix ( bHLH ) factor Ascl1 are the first transcription factors that appear upon initiation of differentiation of sympathoadrenal precursors [13]–[15] . Phox2a is expressed downstream of both Phoxb and Ascl1 [13] , [16] , [17] but its role in the initial stages of the sympathetic lineage remains undefined . Further neuronal differentiation occurs under the influence of the transcription factors Hand2 [18] , [19] , Gata2/3 [20] and Tfap2a [21] , [22] , which interact in a complex regulatory network to ultimately induce the expression of tyrosine hydroxylase ( Th ) and dopamine beta hydroxylase ( Dbh ) - two enzymes required for catecholamine production that serve as markers of terminal sympathetic neuronal differentiation [23]–[25] Extensive work in murine and avian models has established Phox2b as a key regulator of autonomic neuron development , as its complete absence leads to embryonic lethality in mice due to the failure of sympathetic nervous system formation [15] , [26] . Phox2b is first expressed in the murine peripheral sympathetic primordial at the dorsal aorta at E10 . 5 [13] , [27] . By E13 . 5 , it is expressed in all sympathetic progenitor cells including those of the superior cervical ganglia , paravertebral and pelvic ganglia and the adrenal medulla [27] , and expression continues into the early postnatal period [28] . As the ganglia/adrenal medullae mature and the sympathoblasts differentiate , TH- and DBH-expression increases while PHOX2B expression decreases , so that by P28 , only 12% of TH-positive sympathetic neurons have PHOX2B expression ( compared to approximately 60% at P4 ) [28] . Hence , Phox2b is downregulated during terminal neuronal differentiation [28] . Phox2b also has growth inhibitory effects , as its overexpression promotes cell cycle exit and inhibits the proliferation of cultured sympathetic neurons [29] , [30] . Most of the activity of the Phox2b promoter depends on an autoregulatory loop enabling the gene to regulate its own activity and that of other genes [31] . The human PHOX2B gene maps to chromosome 4p13 and consists of three exons encoding a highly conserved 314–amino-acid protein with two polyalanine repeats of 9 and 20 residues C-terminal to the homeodomain . Unlike the case in CCHS , which is defined largely by polyalanine repeat expansion mutations that lead to expansion of the second polyalanine tract [8] , [32] , [33] , germline PHOX2B mutations associated with neuroblastoma tend to be ( i ) missense alterations in highly conserved regions [4] , [7] , [33] or ( ii ) mutations that result in a frameshift , giving rise to an altered or truncated protein lacking the second polyalanine motif [5] , [8] , [33] , [34] . More recently , whole-allele deletions resulting in the reduction of the protein have been reported [35] . Still , the mechanisms by which these different classes of alterations predispose individuals to tumor development in the PSNS remain somewhat unclear . Partial loss of function with the preserved ability to suppress cellular proliferation but not to promote differentiation [36] , complete loss of function due to functional haploinsufficiency [37] and both dominant-negative and gain-of-function effects [30] , [34] , [38] have all been proposed as contributors to neuroblastoma predisposition . In the single in vivo study to date , heterozygous insertion of two frameshift variants , 931del5 and 693del8 , in the mouse Phox2b locus resulted in impaired proliferation of sympathetic ganglion progenitors and biased differentiation towards the glial lineage , leading the authors to conclude that these mutants exhibited both dominant-negative and gain-of-function effects [38] . In this study we sought to analyze the effects of aberrant PHOX2B expression on sympathetic neuron development for each of the major classes of neuroblastoma-associated mutants . We selected the zebrafish model for this purpose because its development occurs ex utero , and the embryos survive considerably longer than mouse embryos , permitting analysis of the PSNS at later stages of sympathetic neuron differentiation and maintenance [39] . We show here that allelic PHOX2B deletion , modeled by morpholino ( MO ) knockdown , leads to a decrease in sympathetic neuronal differentiation in the PSNS . A similar loss of differentiation was observed upon overexpression of a neuroblastoma-linked truncation mutation ( K155X ) [8] and a frameshift mutation ( 676delG ) [5] in the presence of endogenous phox2b , indicating that these variants function in a dominant-negative manner . By contrast , the R100L missense mutation [7] lacked any discernible effect on sympathetic ganglion development in the zebrafish embryo . We demonstrate further that the decrease in terminal differentiation was associated with an increased number of undifferentiated sympathetic neuronal precursors that were resistant to the effects of retinoic acid ( RA ) , and generated a pool of developmentally arrested cells that could serve as targets for future transforming events .
To establish the zebrafish as a model for studying PHOX2B function in PSNS development , we analyzed expression of the phox2b gene in the superior cervical ganglion ( SCG ) . The SCG is the earliest sympathetic ganglion to develop , starting at about 36 hours postfertilization ( hpf ) and progressing in size until about 4 days postfertilization ( dpf ) ( Figure 1A–1D ) [18] , [21] , [39] , [40] . SCG cells are located ventral to the notochord between somites 1 and 4 and are easily visualized as several smaller cell aggregates that progressively increase in size , ultimately forming two separate ganglia in an hourglass shape . phox2b expression is first seen in the SCG cells at 36 hpf extending caudally in two irregular parallel rows to form the primary sympathetic ganglion chain to somite 11 ( Figure 1A–1D ) . It remains robust in the SCG until 72 hpf ( Figure 1C ) when it gradually begins to be replaced by increasing numbers of cells expressing dbh and th , markers of terminal differentiation ( Figures 1E–1F ) , whose expression is first seen at about 48hpf . This expression pattern , especially in the peripheral sympathetic ganglia , and the close sequence homology with human PHOX2B [41] ( Figure S1 ) support the use of the zebrafish to model PHOX2B function in the PSNS . To study the consequences of allelic PHOX2B deletion in patients with neuroblastoma [35] , we performed morpholino ( MO ) knockdown of zebrafish phox2b , using two non-overlapping antisense oligonucleotide sequences targeted to the phox2b gene: a translation-blocking MO ( MOATG ) and a splice-blocking MO ( MOsplice ) directed to the second exon/intron splice junction ( Figure S2A ) . phox2b knockdown was confirmed by immunoblotting in the case of the ATG MO ( which showed ∼70–80% knockdown ) and RT-PCR for the splice MO ( Figure S2B ) . Experiments were performed in both wild-type and p53 mutant embryos ( tp53M214K/M214K ) [42] to account for potential nonspecific effects associated with some MO injections , with similar results obtained in both backgrounds ( Figure S2C ) . MO knockdown of phox2b led to a marked reduction in th and dbh expression in the SCG at 3 dpf , as compared with mismatched control MO-injected ( MOMM ) or uninjected wild-type ( WT ) siblings ( Figures S3A , S3B , S3D , S3E , S3G , S3H ) . This phenotype was consistent with both the ATG and the splice MOs . To ensure that the decrease in th and dbh was not secondary to a general delay in development due to MO injection , we examined the SCG at 4 dpf ( Figures 2A , 2B , 2D , 2E , 2G , 2H ) and later ( 5 dpf; data not shown ) , again noting a decrease in the expression of these genes . To confirm that the phenotype was specific to phox2b , we rescued the th- and dbh-expressing cells by coexpressing human PHOX2B mRNA with both MOs , which led to an increase in th and dbh expression in the SCG ( Figures 2C , 2E; 2G , 2H; Figures S3C , S3F; 3G , 3H ) . These results indicate that PHOX2B is necessary and sufficient for the terminal differentiation of sympathetic neuronal precursors . We next examined the effects of three distinct neuroblastoma-associated PHOX2B mutations on PSNS development ( Figure S4 , Figure 3 ) . Overexpression of the 676delG frameshift [5] and K155X truncation [8] variants led to a significant decrease in the expression of th ( Figure 3D , 3E , 3M ) and dbh in the SCG ( Figure 3J , 3K , 3N ) as compared to that in control ( Figure 3A , 3G ) and WT ( Figure 3B , 3H ) phox2b RNA-injected animals , but had a similar effect to MO knockdown ( Figure 3F , 3L ) . By contrast , overexpression of the R100L homeodomain missense mutation [7] did not lead to a discernible change in the expression of either th or dbh in the SCG ( Figure 3C , 3I ) . To mimic the heterozygous situation seen in patients with PHOX2B mutations , we repeated these experiments in the setting of phox2b MO knockdown . In this context , overexpression of K155X and 676delG led to an even more striking reduction and , in some embryos , to an almost complete absence of th and dbh expression in the SCG ( Figure 3O , 3P ) . These results suggest that the block in differentiation imposed by the 676delG and K155X mutations cannot be rescued by the expression of endogenous wild-type phox2b; rather , these variants appear to function dominant-negatively . PHOX2B complements exogenous differentiating agents such as retinoic acid ( RA ) by promoting cellular differentiation in vitro [36] . To determine whether phox2b loss might obviate the induction of differentiation by RA , we treated control and phox2b MO embryos with various concentrations of 13-cis-retinoic acid , which is commonly used in the treatment of patients with neuroblastoma ( Figure 4A–4F ) . Control-injected embryos showed an increase in th expression in the SCG after RA treatment ( Figure 4A–4C , 4G ) , which was not apparent in the SCG of the phox2b morphant embryos ( Figure 4D–4F , 4G ) . A similar impairment in RA-induced differentiation was observed after expression of the 676delG variant and , to a lesser extent , the K155X variant ( Figure 4H–4P ) . Similar effects in dbh expression were also seen ( data not shown ) . Together , these findings reinforce the strict requirement for Phox2b in the differentiation of sympathetic neuronal progenitors and suggest that loss of this transcription factor cannot be overcome with RA treatment . To assess the impact of decreased phox2b function on the transcription factors that mediate noradrenergic differentiation in the zebrafish , we analyzed their expression following MO knockdown . We observed that mRNA expression of phox2b itself was markedly increased despite the decrease in Phox2b protein induced by MO knockdown ( Figure 5A , 5B; Figure S5A , S5B ) . During normal development in the zebrafish , phox2b expression is first seen in sympathetic ganglia precursors at 36 hpf and , as the cells undergo differentiation , decreases to lower levels by 4 dpf ( this study ) , corresponding to E10 . 5 to E13 in mice [6] . This increase in phox2b RNA expression on abrogation of the protein is consistent with studies that Phox2b regulates its own expression [31] . We also noted that expression of the zebrafish ortholog of ASCL1 , ascl1 , was strikingly increased in the SCG in phox2b morphants ( Figure 5C , 5D; Figure S5C , S5D ) . ascl1 is expressed only transiently in the SCG with maximal expression at 48 hpf , decreasing to less than 10% in 3-dpf embryos [18] . The increased expression of ascl1 transcripts in the SCG of phox2b morphants suggests that Phox2b negatively regulates ascl1 transcription as well . Similar to the effect seen with phox2b MO knockdown , overexpression of the neuroblastoma-associated variants also led to increased ascl1 expression in the SCG , with the 676delG and K155X mutants inducing higher ascl1 expression than did R100L ( Figure 5E–5I ) . Phox2a , a homologue of Phox2b , that is expressed downstream of both Ascl1 and Phoxb during murine sympathetic neural development [13] , [17] was unchanged in phox2b morphant embryos ( Figure 5J , 5K ) . Finally , expression of hand2 [19] , [43] , gata3 [20] , [44] and tfap2a [21] , [22] , three other transcriptional regulators implicated in the control of sympathetic neuronal differentiation [10] , [18] was decreased on phox2b knockdown ( Figure S5E–J ) . Overexpression of the 676delG and K155X variants also led to a decrease in gata3 and tfap2a expression ( Figure S5K–P ) . Together , these results suggest that these genes are regulated by phox2b during sympathetic neuronal differentiation and are affected by perturbations in its function . We surmised that reduced expression of the th and dbh noradrenergic differentiation markers in the SCG of phox2b morphants might reflect either apoptosis of terminally differentiated sympathetic neurons or the failure of precursors to differentiate . There was no evidence of an increase in apoptosis in the SCG as determined by acridine orange staining of phox2b MO-injected embryos ( data not shown ) . However , simultaneous analysis of th and phox2b expression by double labeling of the phox2b morphants showed an increase in phox2b expression with a concomitant decrease in th expression in the SCG compared to stage-matched control MO-injected embryos ( Figures 6A , 6B ) . Analysis of these embryos using qRT-PCR confirmed the presence of significantly increased phox2b expression in the face of decreased th ( Figure 6E ) . The same results were obtained when double labeling was performed with dbh and phox2b , the decrease in dbh-expressing cells being accompanied by increased expression of phox2b in morphants compared to controls ( Figures 6C , 6D ) . We were unable to test whether this was also true in embryos in which PHOX2B variants were overexpressed because the high level of sequence homology between human and zebrafish phox2b made it difficult to interpret the results . To confirm that these phox2b- and ascl1-expressing SCG cells were arrested early in differentiation , we analyzed the expression of elavl3 ( HuC ) , the well-characterized , evolutionarily conserved neural differentiation marker that is normally only detected in mature postmitotic neurons and has been used to demonstrate neuronal differentiation in the zebrafish [41] , [45]–[47] ( Figures 6F–6K ) . We observed a decrease in elavl3 expression in the phox2b morphants compared to controls , with the extent of reduction comparable to that of th and dbh , suggesting that the SCG is populated mainly by undifferentiated cells . Overexpression of the 676delG and K155X mutants , but not R100L , also led to a decrease in elavl3 expression in the SCG of these embryos ( Figures 6H–K ) . Together , these data suggest that a decrease in Phox2b levels , either through MO knockdown or overexpression of dominant-negative mutations , blocks sympathetic cells at a progenitor stage marked by high phox2b and ascl1 expression . Such cells cannot proceed to the next developmental stage , involving hand2 , gata3 and tfap2a , and therefore cannot undergo terminal PSNS differentiation characterized by the expression of th and dbh . During noradrenergic development in the chick embryo , Phox2b and Ascl1 are expressed together in response to BMP signaling [44] , [48]; in fact , elimination of Ascl1 has also been reported to cause impaired sympathetic differentiation [14] , [16] . To determine whether ascl1 is as critical as phox2b to PSNS differentiation in the zebrafish , we studied the effects of ascl1 knockdown using a translation-blocking MO ( Figure S6A; Figure 7A , 7B , 7C , 7E ) . Abrogation of ascl1 alone led to only a marginal decrease in th and dbh expression in the SCG ( Figure 7B , 7E , 7G ) , while a striking decrease in both th and dbh expression occurred with simultaneous knockdown of both ascl1 and phox2b , similar to the result with phox2b knockdown alone ( Figure 7C , 7F , 7G ) . Moreover , in contrast to the outcome of phox2b knockdown , expression of hand2 , gata3 and tfap2a was not affected in the ascl1 morphants ( Figure S6B–G ) . These findings identify Phox2b as the central driver of terminal differentiation of sympathetic neurons in the zebrafish model . Observations in murine models have revealed that PHOX2B and ASCL1 cross regulate each other [6] , [14] , [16] , [49] . The fact that ascl1 expression was significantly increased in the phox2b morphants suggested a regulatory effect of phox2b on ascl1 ( Figure 5C , 5D ) . To determine if Ascl1 had the same effect on phox2b in the zebrafish model , we studied the effects of ascl1 knockdown on phox2b expression . While causing a decrease in its own expression at both the RNA and protein levels ( Figure 7H , 7I , Figure S6A ) , knockdown of ascl1 had no effect on phox2b expression ( Figure 7K , 7L ) . Similarly , overexpression of ascl RNA also did not affect phox2b expression ( Figure S6H , S6I ) . However , knockdown of both genes simultaneously led to a striking increase in ascl1 ( Figure 7J ) and phox2b ( Figure 7M ) mRNA expression in the SCG . These results suggest that although Phox2b is capable of regulating ascl1 in the zebrafish model , the latter does not have any appreciable impact on phox2b . Rather , Ascl1 appears to primarily regulate phox2a . Unlike knockdown of phox2b , which did not affect phox2a expression , we observed a significant reduction in phox2a expression in the SCG following ascl1 knockdown ( Figure 7N , 7O ) . These results are consistent with the finding that Ascl1 is required for the expression of Phox2a in sympathetic and parasympathetic murine ganglionic anlagen [14] , [16] , with Phox2b assuming a less important role .
In this study , we relied on a zebrafish model of PSNS development to demonstrate that a reduction in phox2b expression due to MO knockdown or overexpression of the neuroblastoma-associated 676delG frameshift and K155X truncation variants of PHOX2B inhibits the terminal differentiation of sympathetic neuron progenitors ( Figure 8 ) . Our findings indicate that intact Phox2b function is essential for the normal development of cells in the sympathetic ganglia , and that in the context of phox2b deficiency , differentiation cannot be induced by exogenous agents . In contrast to the remainder of the nervous system , where the onset of neuronal differentiation is coupled with cell cycle exit , an increase in the number of sympathetic neurons during development reflects the proliferation of differentiated sympathetic neurons rather than the proliferation and subsequent differentiation of neuronal progenitors [50] , [51] . Indeed , studies in the chick embryo have revealed that the vast majority of postmitotic sympathetic neurons are generated through the proliferation of already differentiated neurons [50]–[53] . An alteration in Phox2b function in the zebrafish , by either MO knockdown or overexpression of certain neuroblastoma-associated variants , led to a block in the differentiation of sympathetic progenitor cells . First , there was a large increase in phox2b transcript-expressing cells at the expense of th- and dbh-expressing cells in the SCG ( Figure 6 ) . Second , these phox2b-expressing cells were unable to proceed to subsequent differentiation steps , even when stimulated with RA , whose signaling can induce noradrenergic differentiation in the zebrafish ( Figure 4 ) [21] . Third , in addition to decreased th and dbh expression , the SCG cells also showed reduced expression of the mature neuronal marker elavl3 ( Figure 6 ) . In our study , specific types of PHOX2B variants exhibited distinct effects on sympathetic neuronal differentiation , with the R100L missense mutation lacking an effect on differentiation , while the 676delG frameshift and K155X truncation mutations exhibited impaired differentiation , an effect that was all the more prominent with simultaneous MO knockdown . Findings similar to ours were reported by Reiff et al ( 2010 ) on overexpression of the K155X mutation in primary chick sympathetic neuron cultures , although in this system the 676delG variant lacked any apparent effect on the expression of terminal differentiation markers [30] . In agreement with our data , another group reported resistance to differentiation with RA in human neuroblastoma cells engineered to express the 676delG variant ( in the presence of endogenous PHOX2B ) [36] . The R100L homeodomain variant , on the other hand , did not appear to affect sympathetic neuronal differentiation in the zebrafish , similar to findings in the above mentioned study in avian sympathetic neurons , where this mutant elicited increased cell proliferation only [30] . The basis for the varied effects of distinct mutations on sympathetic neuronal differentiation is not entirely clear . To some extent , they reflect intrinsic differences between the in vitro and in vivo models in which the variants were tested , as well as different degrees of forced or blocked expression in target cells . Ultimately , however , the impact of PHOX2B mutants on the differentiation of immature sympathetic neurons depends on the protein structures that are modified . The R100L missense variant , for example , is defined by a minimal change in its homeodomain , whereas both 676delG and K155X possess modifications due to an altered reading frame or deletion of critical coding sequences within the C-terminus , respectively , either of which can lead to the disruption of essential protein-protein interactions or , in some instances , to novel interactions not seen with the WT protein . Indeed , misfolding and oligomerization have been demonstrated with frameshift mutations of PHOX2B and cellular mislocalization and cytoplasmic aggregation with truncated variants [54] . The 693del8 frameshift mutant , for example , is thought to interact with proteins that do not bind to the WT protein [38] . Similarly , we have noted that the 676delG and K155X variants do not bind to certain proteins recognized by WT PHOX2B , resulting in impaired neuroblastoma cell differentiation ( George et al , unpublished observations ) . Thus , it is not surprising that the K155X truncation mutation , unlike the R100L missense change , generates a protein that not only stimulates sympathetic neuron proliferation [30] , but also has a dominant–negative inhibitory effect on cell differentiation . Furthermore , since both frameshift and truncation changes in the PHOX2B gene appear to generate stable proteins that lack the transactivation potential of WT PHOX2B , [30] , [36] , [37] , [55] their contribution to neuroblastoma predisposition could be twofold . By interacting with different modifier proteins , these mutants could preserve the ability of PHOX2B to suppress cellular proliferation while abolishing its pro-differentiation regulatory effects . At the same time , as has been shown previously [30] these transactivation–impaired variants likely compete with intact PHOX2B for critical promoter sequences on neuronal differentiation-linked target genes , resulting in a dominant-negative repressive effect on their expression . This hypothesis is supported by the failure of the WT phox2b to rescue the arrested differentiation of sympathetic neuronal progenitors expressing either the 676delG or K155X variants . The gene dosage of PHOX2B also appears to have played a role in determining the disease phenotype in our study . Indeed , fish with MO knockdown of phox2b but no mutations or deletions consistently showed arrested differentiation of sympathetic neuronal progenitors . This result would account for a neuroblastoma case recently reported by Jennings et al [35] in which a heterozygous deletion of PHOX2B was associated with both CCHS and a neural crest tumor . Thus , PHOX2B deficiency due to whole-allele deletion should be considered another mechanism whereby individuals might acquire a predisposition to neuroblastoma . The precise oncogenic mechanism of PHOX2B deficiency and its associated phenotypes will require additional study , but it seems likely that dosage reduction beyond 50% ( sub-haploinsufficiency ) may be required , as mice that were haploinsufficient for Phox2b did not develop tumors [56] . Neuroblastoma development in TH-MYCN transgenic mice begins with hyperplastic lesions in early postnatal sympathetic ganglia that are composed predominantly of Phox2b-positive but Th-negative neuronal progenitors [28] . This indicates a central role for aberrant PHOX2B regulation of immature sympathetic neurons in neuroblastoma predisposition . Our studies suggest a model ( Figure 8 ) in which aberrant PHOX2B function through either allelic deletion or dominant-negative mutations promote the same endpoint: impaired differentiation of sympathetic neuronal progenitor cells while promoting the expansion of a population of undifferentiated cells likely to be susceptible to so-called second “hits” such as MYCN amplification .
All experiments involving zebrafish conformed to the regulatory standards and guidelines of the Dana-Farber Cancer Institute ( DFCI ) and the Brigham and Women's Hospital ( BWH ) Institutional Animal Care and Use Committee . Zebrafish were maintained at the DFCI and BWH zebrafish facilities under standard conditions [57] . Embryos were raised in E3 medium supplemented with 0 . 003% 1-phenyl-2-thiourea ( PTU ) at 28 . 5°C to allow visualization of internal structures . The p53−/− mutant allele tp53M214K/M214K was kindly provided by A . Thomas Look [42] . tp53M214K/M214K and wild-type embryos of the AB strain were obtained by natural group mating . The embryos were staged according to established morphological criteria [57] . Morpholinos ( MOs ) were obtained from GeneTools , LLC ( Philomath , OR ) based on the published GenBank sequences for phox2b and ascl1 ( phox2b P2BT2 splice donor: 5′-AAGTAAGCGGAGAATGTCCCACCTG; mismatched phox2b P2BT2 splice donor: AcTAAcCGGAcAATcTCCgACCTG; phox2b ATG: 5′-TATACATTGAAAAGGCTCAGTGGAG; mismatched phox2b ATG: TATAgATTcAAAAccCTCAcTGGAG; ascl1 ATG: 5′- CCATCTTGGCGGTGATGTCCATTTC; mismatched ascl1 ATG: CCATgTTGGCcGTcATcTCgATTTC ) . For knockdown of phox2b , two nonoverlapping MOs were used ( ATGK1 , designed to target the start site; and P2BT2 , designed to block splicing at the phox2b exon 2-intron 2 boundary ) . Embryos at the one to two-cell stages were injected with up to 4 ng of MO diluted with 1× Danieau's solution with Phenol Red . Injected embryos were raised to 12 , 24 , 36 , 72 , 96 and 120 hpf and processed for in situ hybdridization . Generalized nonspecific necrosis in MO-injected and mismatched control MO-injected AB embryos was alleviated using the tp53M214K/M214K strain [42] . MO concentrations were optimized so that the lowest amount necessary to retain the normal phenotype while demonstrating effects of absent phox2b expression was used . The observed phenotypes were consistent between the ATG and splice MO . Care was taken to stage-match the embryos as much as possible with mismatched MO-injected controls so as to eliminate phenotypes due to developmental delay . Human PHOX2B ( a generous gift from K-S Kim ) and the PHOX2B mutant 676delG ( kindly provided by J . Maris ) were subcloned into the pCS2+ vector . Site directed mutagenesis was used to generate PHOX2B mutant constructs R100L and K155X using the Quickchange II Site-directed mutagenesis kit ( Stratagene ) . ascl1 was amplified from a whole embryo cDNA library and cloned into pCS2+ . The SP6 Message machine kit ( Ambion ) was used to transcribe synthetic capped RNA . Embryos were injected with 50–100 pg of mRNA at the one- to two-cell stages . Injected embryos were raised at different time points ranging from 12 hpf to 4 dpf and fixed with 4% paraformaldehyde for in situ hybridization . The following antisense RNA probes were generous gifts: phox2b and phox2a ( S . Guo ) , dbh and elavl3 ( J-S Lee ) . Other probes ( th , ascl1 , tfap2a , hand2 , gata3 ) were generated by amplifying the Danio rerio ORFs from a whole embryo cDNA library using PCR primers based on published GeneBank sequences . The PCR products were subcloned into the EcoRI/XhoI sites of pCS2+ vector . Following sequence verification , antisense riboprobes were generated by in vitro transcription with DIG RNA labeling kit Sp6/T7 ( Roche ) . Embryos at different developmental stages were collected and fixed in 4% ( w/v ) paraformaldehyde ( Sigma ) . Whole-mount in situ hybridization was performed as described [58] . Images of zebrafish embryos were taken using an Olympus SZX12 microscope and a digital camera . Intensity measurements were performed using Image-Pro Plus ( IPP ) software ( MediaCybernetics , PA ) . Identical regions were selected using the rectangle tool set to a constant area . Mean optical density ( MOD ) and total per area ( TPA ) in boxed areas were used for pixel intensity . Differences between two groups were analyzed using Student's t-test . At 2 or 3 dpf embryos injected with a control MO and phox2b MO , were subjected to 13-cis retinoic acid treatment for 24 hrs at three concentrations 1 nM , 10 nM , and 100 nM diluted in PTU egg water . DMSO was used as a control . Embryos were then fixed and gene expression was analyzed using in-situ hybridization . Embryos were homogenized in TRIzol reagent by passing through a 22G needle . Total RNA was extracted as described in the manufacturer's instructions ( Invitrogen ) . Primer pairs encompassing the full-length phox2b gene were designed and used to amplify phox2b in the MO-injected embryos with the Superscript First strand synthesis kit ( Invitrogen ) . 50 pg of 1st strand cDNA was used to amplify phox2b using the Expand High Fidelity Plus PCR system ( Roche ) . Immunoblotting was performed according to standard methods . The following antibodies were used: anti-PHOX2B ( Santa Cruz , N14 cat# sc-48627 ) , anti-ASCL1 ( Santa Cruz , sc-28688 ASCL1-H56 ) . Anti-beta tubulin antibody ( Abcam 6046-100 ) was used as a loading control . | Neuroblastoma , a tumor of the peripheral sympathetic nervous system , is the most common cancer diagnosed in infancy . Although most cases arise sporadically , familial predisposition also occurs in association with mutations in a single copy of the PHOX2B gene , a “master regulator” of sympathetic neuronal development . The exact mechanisms by which these mutations increase susceptibility to neuroblastoma are unclear , primarily because of the paucity of optimal models in which to study very early development of the sympathetic nervous system . We took advantage of the ex vivo development and transparent nature of zebrafish embryos to study the roles of both normal and mutated PHOX2B in development of the sympathetic nervous system . We present data indicating that aberrant PHOX2B expression causes an arrest in the normal maturation of sympathetic neurons , leading to immature cells that are resistant to drug-induced differentiation . Indeed , we demonstrate that phox2b gene “dosage” is important for normal differentiation of sympathetic neurons in the zebrafish and suggest that the population of immature cells resulting from a decreased dosage of this pivotal factor may be susceptible to secondary mutations that could ultimately lead to neuroblastoma . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"oncology",
"medicine",
"neurological",
"tumors",
"developmental",
"biology",
"developmental",
"neuroscience",
"neurobiology",
"of",
"disease",
"and",
"regeneration",
"cancer",
"genetics",
"pediatric",
"oncology",
"genetics",
"biology",
"cancers",
"and",
"neoplasms",
"gen... | 2013 | Distinct Neuroblastoma-associated Alterations of PHOX2B Impair Sympathetic Neuronal Differentiation in Zebrafish Models |
Apicomplexa are obligate intracellular pathogens that have fine-tuned their proliferative strategies to match a large variety of host cells . A critical aspect of this adaptation is a flexible cell cycle that remains poorly understood at the mechanistic level . Here we describe a forward genetic dissection of the apicomplexan cell cycle using the Toxoplasma model . By high-throughput screening , we have isolated 165 temperature sensitive parasite growth mutants . Phenotypic analysis of these mutants suggests regulated progression through the parasite cell cycle with defined phases and checkpoints . These analyses also highlight the critical importance of the peculiar intranuclear spindle as the physical hub of cell cycle regulation . To link these phenotypes to parasite genes , we have developed a robust complementation system based on a genomic cosmid library . Using this approach , we have so far complemented 22 temperature sensitive mutants and identified 18 candidate loci , eight of which were independently confirmed using a set of sequenced and arrayed cosmids . For three of these loci we have identified the mutant allele . The genes identified include regulators of spindle formation , nuclear trafficking , and protein degradation . The genetic approach described here should be widely applicable to numerous essential aspects of parasite biology .
Apicomplexans are highly successful protozoan parasites infecting a tremendous variety of vertebrate and invertebrate animals . In humans they are responsible for several important diseases , including malaria , toxoplasmosis , and cryptosporidiosis . A key to their success is their adaptation to a unique intracellular niche , which allows them ready access to nutrients while sheltering them from the immune system . Asexual parasite replication is restricted to this intracellular part of the life cycle , and intermediate stages of intracellular replication in many species lack the machinery to infect new host cells . It is therefore critical for the parasite to time its cell division and the formation of invasive forms to coincide precisely with host cell egress . To adapt to a variety of specific host cell niches , apicomplexa have developed several specialized cell division modes . These division modes are based on a ‘flexible' cell and division cycle program that can actively coordinate DNA synthesis and chromosome segregation , while at the same time suspending nuclear division and/or cytokinesis until the last step of the replication cycle ( see [1 , 2] for recent detailed reviews of apicomplexan cell division and cell cycle control ) . Thus , re-initiation ( s ) of DNA synthesis prior to the completion of cytokinesis occurs naturally in these parasites . The molecular mechanisms that count the rounds of DNA synthesis and provide the proper timing of parasite budding remain one of the compelling mysteries of these parasites . Database mining for factors commonly associated with eukaryotic cell cycle control has identified an extensive set of candidate regulatory proteins in apicomplexan parasites ( e . g . cyclins , CDKs , MAPKs [3–6] ) . While these findings predict that cell cycle checkpoints exist in Apicomplexa , they do not provide information about where checkpoints function or how these controls operate to safeguard the diverse strategies utilized by these parasites . Here we describe the genetic analysis of the apicomplexan cell division machinery in Toxoplasma gondii . While the simple binary division of Toxoplasma tachyzoites ( also termed endodyogeny ) offers an attractive model system , we expect these studies to apply broadly to the replication of other pathogens in this phylum , such as Plasmodium , Eimeria , and Cryptosporidium , where our knowledge of the parasite cell cycle is equally deficient . Using chemical mutagenesis and a high-throughput replica assay , we have isolated a large collection of temperature sensitive ( ts ) parasite mutants . Our phenotypic analyses map these mutants to specific steps of the parasite cell and division cycle . To identify the underlying genes we have developed a robust complementation model employing cosmid transformation . This approach allowed us to link mutant phenotypes to specific point mutations in parasite genes .
A pool of conditional growth mutants was established by chemical mutagenesis using N-nitroso-N-ethylurea ( ENU ) , an agent used successfully in the past to generate point mutations in the T . gondii genome [7–10] . ENU was applied at a dose inducing 60–70% parasite killing ( measured through plaque assay; results not shown ) . We estimate this dose to induce 10–100 mutations per genome based on the incidence of mutations in hypoxanthine-xanthine-guanosine phosphoribosyltransferase ( HXGPRT [11 , 12] ) , which we measured by following the emergence of resistance to the HXGPRT activated prodrug 6-thioxanthine ( 10−5 in our experiments , also see [7] ) . To enable high-throughput screening we developed a well plate replica assay ( see Figure 1 for a schematic outline of the strategy ) . Following mutagenesis , parasites were immediately cloned into 384 ( or 96 ) well plates seeded with HFF cells to avoid competition with wild type parasites and allowed to expand at the permissive temperature ( 34°C or 35°C ) . To identify mutants , plates were duplicated using a pin-tool transferring 5 ( or 20 ) μl parasite suspension into two new plates . One plate was kept at the permissive temperature , whereas the replicate was placed at 40°C ( restrictive temperature ) . Two approaches were applied to detect temperature sensitivity: visual microscopic inspection of wells ( mutagenizing parent strain RH/hxgprt− ) and measurement of fluorescence ( mutagenizing the autofluorescent reporter strain 2F-1-YFP2 [13] ) . Fluorescence was measured after 4 days ( 40°C ) and 7 days ( 35°C ) . The values were normalized against the 2F-1-YFP2-parent line included in each plate , and corresponding wells were compared for differential growth at permissive and restrictive temperatures using an automated script ( a fluorescence increase below 20% was scored as a growth phenotype ) . Parasites from wells exhibiting growth at 35°C but not at 40°C were expanded , and temperature sensitivity was re-confirmed ( see Figure 1B for examples; mutants with an identifier that starts with a letter were obtained through the fluorescence screen , those that start with a number through the visual screen ) . In total we have identified 165 ts mutants from ∼60 , 000 clones produced ( see Table S1 ) . Approximately 5% of the primary clonal isolates selected in the screens were confirmed by secondary analysis to show conditional growth arrest at the restrictive temperature ( 165/2960 ) . False positives arose equally using either detection method and were due to the liberal selection for potential growth mutants and also as the result of replica-pin transfer failures ( ∼5% ) . The overall frequency of confirmed ts mutants generated by our combined screens is lower ( 0 . 26% ) than the value obtained in an earlier pilot screen ( 1 . 1% , [14] ) , which was much smaller in scale ( ∼3 , 600 total ENU clones screened ) and employed less stringent criteria for validating conditional growth . The ts mutants produced here display a lower reversion to wild type growth than our earlier study with >85% of clones in the current screen having reversion frequencies that are <10−6 and many isolates revert at <10−7 ( see Table S2 ) . The ts mutants generated by this and our earlier study [14] display almost exclusively conditional-lethality . However , a few examples of mutants where viability is retained following temperature shift were observed; as noted in other eukaryotic models [15 , 16] these are mostly G1 mutants ( see below ) . Mutants were analyzed for uniform population changes in DNA content measured by flow cytometry ( FACS ) . Phenotypes were further characterized by immunofluorescence assays ( IFAs ) identifying distinctive cellular and nuclear morphologies that developed at the restrictive temperature ( see Table S2 for detailed individual descriptions of all mutants examined ) . Specific antibody reagents used in IFA analyses were directed against centrin as a marker for the number and position of centrosomes [17–19] , membrane occupation recognition nexus ( MORN ) 1 , a marker for spindle morphology and budding [19 , 20] , TgPCNA1 , an essential element of the nuclear DNA replication complex [21 , 22] , and inner membrane complex ( IMC ) 1 and 3 , components of the membrane skeleton that served as a budding marker [23–25] . The majority of parasite mutants examined had phenotypes characteristic of growth arrest in a specific cell cycle phase ( i . e . >75% of the parasite population examined by FACS or IFA show a similar phenotype , see Figures 2 , 3 , and 4 for representative examples ) . In comparison to similar efforts in Saccharomyces [26 , 27] , fewer general growth mutants were produced by our screens ( the overall yield of ts mutants is also lower ) . However , several non-cell cycle mutants were identified including e . g . F-P2 , a mutant with normal intracellular development but a severe invasion and egress defect ( M . J . Gubbels and B . Striepen , unpublished data ) . Based on their shared terminal phenotypes , mutants were readily classified into groups . These groups express defects in mechanisms active across the full spectrum of events in parasite replication . As expected , cell cycle mutants were isolated that possess a dominant haploid DNA content ( 1N ) at the restricted temperature . The Sytox Green-FACS histograms for mutant 88A5 ( Figure 2B , 34°C versus 40°C ) are representative of this large group of G1 phase mutants . In addition to a 1N DNA content , G1 mutants possessed a single nucleus and had no or few internal daughter forms at the time of growth arrest ( data not shown , see Table S2 ) . Outside this core set of phenotypic features , G1 mutants in our collection express diverse secondary characteristics that include alterations in terminal cell size ( e . g . mutant 73C1 , Figure 3B ) , differences in the number of cell divisions before growth arrest , and variations in intermediate temperature sensitivity . The G1 class is the only group where non-lethal ts mutants were isolated in our screen . Mutants 63H4 and 31F1 stop within a single cell division in the G1 phase when shifted to 40°C and can be held at this temperature for 24 hours without significant loss of viability ( 116% or 92% plaques formed compared to controls maintained constantly at the permissive temperature . Note that most mutants show poor recovery in this assay e . g . 88A5 or 87A10 with 1 or 4% , respectively ) . Two unusual mutants classified in the G1 group display complete loss of nuclear proliferating cell nuclear antigen ( PCNA ) 1 staining at the restrictive temperature ( e . g mutant 124H2 , Figure 3A ) . A second group of ts mutants in the collection arrest upon shift to 40°C with an intermediate DNA content ( >1N but <2N by Sytox Green-FACS ) consistent with S phase arrest . Five ts mutants representing this class ( e . g . 150B8 , Figure 2C ) arrest with a 30% increase over the haploid DNA content at the restrictive temperature; this phenotype is very similar to RHTK+-parasites blocked by thymidine treatment in early S [28] . Arrested cells had large , centrally located nuclei and the DAPI and nuclear PCNA1 staining patterns observed ( Figure 3C ) were consistent with the S phase assignment [22] . Two additional S phase mutants deserve mention here: mutant 150B10 possessed a mid-S phase DNA content at the time of growth arrest ( Figure 2D ) , while mutant 104A4 ( Figure 2E ) showed a unique bimodal distribution of parasites into equal 1N and 1 . 8N subpopulations . Internal budding is a unique feature of apicomplexan replication , and >30% of the initial ts mutants characterized by IFA and FACS display defects in cytokinesis , some of these show simultaneous defects in karyokinesis . Mutant 64D5 is representative of five mutants that have defects in chromosome segregation . A subpopulation with a sub-1N DNA content by FACS ( Figure 2F ) as well as microscopic evidence for zoid formation ( anucleate daughter cell [29] , Figure 3D ) are key characteristics of these mis-segregation mutants . A second group of M-phase mutants shows defects in the early stages of budding . Multiple small IMC1 staining structures formed but failed to develop fully when the parasites were shifted to the restrictive temperature ( Figure 4A ) . Other budding mutants developed defects that interfered with the resolution of the mature daughters from the mother cell . Apparently this did not prevent a second round of cell division from unfolding suggesting cell cycle counting mechanisms were still active in this late budding mutant ( bud-within-bud mutant 7A11 , Figure 4D ) . The early and late budding mutant examples shown here maintained normal numbers of nuclei per parasite ( one or two ) , whereas in other mutants isolated by our screens a catastrophic breakdown of the coordination between cytokinesis and karyokinesis occurred at the restrictive temperature . The three examples of this type of mutant shown here were promiscuous for chromosome re-initiation , which led to the formation of multiple or very large nuclei ( 42D6 and PO-B3 , Figure 4B and 4C , and V-A15 see below ) . In each mutant , the mitotic spindle apparatus appears disorganized based on MORN1 antibody staining ( all three mutants had similar staining as shown for mutant V-A15 below ) . Interestingly , similar abnormal budding and massive DNA over-replication has been observed in parasites where the spindle has been disrupted pharmacologically [29] , suggesting that spindle defects might be a key feature of the uncoupling phenotype expressed by these ts mutants . To identify the genes underlying the mutant phenotypes , we employed phenotypic complementation using a wild-type genomic DNA library . Several T . gondii libraries have been generated for this purpose and some success has been reported [25 , 30–32] . However , in our initial experiments we found that the established cDNA library strategies were not sufficiently robust to provide complementation on a regular basis with the ts mutant pool produced in this study . This is likely due to limitations of cDNA based plasmid libraries with respect to pool redundancy and insert size ( many cell cycle factors are encoded by low abundant mRNAs ) . To establish a complementation model that is independent of gene size and differential transcript levels , we generated a RH genomic DNA cosmid library . T . gondii does not maintain stable episomes precluding a simple shuttle of cosmids between parasite and E . coli to isolate complementing sequences . We therefore adapted a strategy based on cosmid insertion and rescue of a sequence tag ( see Figure 5 for a schematic outline ) . Our library was constructed in ToxoSuperCos , a double cos-site plasmid based on the commercial SuperCos1 construct ( Stratagene ) . However , several features were engineered into ToxoSuperCos to facilitate rescue after insertion . The T . gondii pyrimethamine resistance marker DHFR-TSm2m3 [33] was included to select for stable cosmid integration into the parasite genome preserving the backbone . This is important as plasmid rescue requires a bacterial origin of replication and a drug resistance marker for selection in bacteria ( [34] , see Figure 5 ) . We chose a kanamycin resistance gene , as most parasites used for screening already contain plasmid DNA harboring ampicillin resistance genes ( due to previous genetic engineering to introduce reporters and markers ) . To further facilitate rescue , a polylinker was incorporated providing a broader choice of restriction sites for genomic excision of rescue tags . The resulting library consists of 1 . 25x106 independent cosmid clones providing ∼900-fold genomic coverage . Using pyrimethamine selection and parasite plaque assays , the transformation efficiency was determined to be 0 . 3% ( which is comparable to plasmid based transfection taking the larger size of the cosmid construct into account ) . Using the protocol described in the materials and methods section , 12-fold coverage of the T . gondii genome should be achieved in each transfection experiment . To validate these calculations , we transfected the T . gondii RH-hxgprt- deletion mutant with the cosmid library and selected for complementation by treatment with mycophenolic acid [11] and pyrimethamine . Five out of five electroporations resulted in the isolation of viable and inheritably mycophenolic acid resistant parasites . Clonal lines were established from three independent transfections , and the wild type HXGPRT locus ( absent from the mutant but present in the library ) was detected by PCR and sequencing of the PCR product in ten out of fifteen clones ( Figure S1 ) . Further PCR-based analyses indicated that complementation occurs through heterologous insertion of an additional wild type copy rather than homologous gene replacement of the mutant locus ( even in experiments where no selection for the pyrimethamine marker is applied ) . Having established that the ToxoSuperCos library robustly complements the HXGPRT mutant , we then tested its ability to genetically rescue ts mutants . Here we describe experiments with mutant V-A15 in detail as an example , but note that subsequently numerous additional ts mutants have been successfully complemented with this library ( see below ) . Mutant V-A15 shows tight temperature sensitivity , with a modest growth delay at 35°C , significant inhibition at 37°C and a severe defect at 40°C ( Figure 1B; the reversion frequency of this mutant was measured by plaque assay to be <10−7 ) . IFA and FACS analyses revealed that growth inhibition is due to severe mitotic defects . V-A15 parasites shifted to the restrictive temperature failed to complete mitosis which led to polyploid nuclei and/or chromosome loss ( Figure 6B , 6D , and 6F; note that increase in nuclear size goes along with an increase in cell size ) . While internal daughter buds were readily observed in parasites grown at the permissive temperature by staining with IMC3 ( ∼25% of all vacuoles , Figure 6A and 6C ) , no clear IMC3 structures were formed at the restrictive temperature . Because features of this phenotype are consistent with a defect in the mitotic spindle , we analyzed mutants for the cellular distribution of MORN1 , a marker of the nuclear spindle compartment ( centrocone [20] ) . In parasites grown at the permissive temperature , the centrocone was clearly detected in each nucleus , while under temperature restrictive conditions MORN1 nuclear staining appeared disorganized or was entirely absent from some of the nuclei . To identify the underlying genetic locus , mutant V-A15 was transfected with the ToxoSuperCos library in five independent electroporations , inoculated into confluent HFF cultures and allowed to recover for 24 hrs at 35°C . The flasks were subsequently transferred to 40°C to select for growth restoration ( phenotypic complementation ) , and pyrimethamine was added to select for stable cosmid backbone integration . Stable temperature and drug resistant parasites emerged in four out of five flasks . To identify the complementing cosmid sequence , we used plasmid rescue of a sequence tag ( see Figure 5 for an outline of this strategy ) . Genomic DNA of complemented parasite lines was extracted and digested with SpeI , HindIII or BglII , self-ligated overnight , and then electroporated into E . coli . Kanamycin resistant colonies were recovered for all four complementations using at least two out of the three enzymes used . Plasmid DNA was isolated and the genomic inserts were end-sequenced using the T3 primer . Three out of the four sequence tags ( each from independent complementation experiments ) mapped to a 30 , 000 bp locus on chromosome IX and a region of 6 . 7 kb could be identified that was shared between all three rescued tags ( see Figure 7A; the fourth complement mapped to Chr VIII 1225114 bp minus strand , and was not studied further ) . The overlapping region contains a single predicted gene model ( 80 . m02355 ) encoding a putative NIMA-related kinase with close similarity to Plasmodium falciparum Nek1 [6 , 35] . Rescue of a sequence tag as described above readily identifies a candidate complementing locus , yet only a small portion of the complementing sequence is cloned into plasmid , which is often not sufficient to independently confirm the result by re-complementation . However , the ToxoSuperCos library ( along with a second library constructed by Dan Howe and David Sibley , pSCBle , http://toxomap . wustl . edu/ ) has been end-sequenced in the course of the T . gondii genome project , and the cosmid tiling ( covering essentially the entire genome ) can be viewed through ToxoDB ( http://www . toxodb . org/ ancillary genome browser ) . To test if the locus identified is indeed sufficient to complement the mutation , V-A15 was transfected with cosmid ToxPJ50 ( containing the TgNek1 locus , Figure 7A ) or ToxP932 ( an unrelated control cosmid; Chr VIIb 1 , 484 , 413bp-1 , 521 , 843bp ) . Stable transgenics were established by pyrimethamine selection at the permissive temperature . These parasites were scored for temperature sensitivity by plaque assay ( parasite growth is indicated by host cell lysis resulting in clear plaques in the fibroblast monolayer ) and fluorescence growth assay ( measuring the fluorescence of the yellow fluorescent protein ( YFP ) -YFP transgene expressed by the parasites [13] ) . ToxPJ50 restores robust growth at 40°C , while the control transgenic is indistinguishable from the mutant ( Figure 7B–7G , note that complementation with ToxPJ50 already confers a modest growth advantage at 35°C , panel F ) . The complementation data indicate that the TgNek1 locus is the site of the mutation causing the V-A15 phenotype , or alternatively , that this locus can act as a suppressor . To distinguish these two possibilities , we amplified the locus by PCR from both wild-type parent and mutant V-A15 , transfected the mutant with each allelic gene fragment and scored for complementation by plaque assay . PCR product from wild-type complemented in three out of three independent experiments , while no growth was observed with the V-A15 derived gene fragment ( Figure 8A–8D ) , suggesting that the mutation is localized within this locus . Importantly , complementation with the wild type PCR fragment also restores the FACS DNA profile of parasites grown at 40°C to the typical wild type distribution ( Figure 8F ) . The 7 . 2 kb PCR fragments were sequenced on both strands and compared to the genome sequence as well as each other . A single base pair change distinguishes the sequence of the mutant from RH wild-type . This change of a T to C lies within the Nek1 coding region and changes a cysteine to an arginine in a highly conserved portion of the predicted protein ( Figure 8E shows the mutation along with a short alignment of Nek1 from T . gondii and P . falciparum ) . Taken together , these data indicate that a point mutation in TgNek1 is responsible for the severe mitotic defects observed in this mutant . Encouraged by our success in complementation analysis of mutant tsV-A15 , we have broadened our efforts to ultimately identify the genes affected in all of the 165 ts mutants isolated in this screen . Figure 9 summarizes phenotypic and genetic analyses for the first group of 41 mutants ( 7 additional mutants showed excessive reversion frequencies and were excluded from further analysis ) . While these studies are still ongoing , thus far , we have observed complementation ( indicated by asterisks ) for 22 out of 24 attempted mutants , and a candidate locus has been identified by marker rescue for 18 mutants ( see Table 1 ) . For 8 mutants the locus has been independently confirmed by successful re-complementation using the respective arrayed cosmid , and for three genes the mutant allele has been identified by sequencing as described above for V-A15 ( underlined in Figure 9; this data is summarized in detail in Table 1 ) . The genes identified through our mutant analyses encode proteins with potential functions in a wide array of parasite cell cycle mechanisms . Eleven of the 14 T . gondii genes summarized in Table 1 have clear homologs in other apicomplexan genomes as identified by OrthoMCL . In several cases , they represent orthologs of known cell cycle factors from other eukaryotes . Orthologs of the NimA-related kinase that complements V-A15 have well described roles in centrosome biology and mitotic entry [35–38] . Other examples include the Toxoplasma protein encoded by 27 . m00873 , which appears to belong to the Sac3/GNAP family . The SAC3 gene product in yeast is a nuclear factor that is required for mitotic progression . Defects in this gene cause errors in yeast budding and mitotic delay [39] , which compares with the failure of mutant 118G4 to properly progress into mitosis and budding . An AAA-ATPase ( 44 . m0215 ) is present in the locus complementing S phase mutant 104A4 . AAA-ATPases fold and unfold proteins and often act as gatekeepers of protein degradation [40] . Mutations in AAA-ATPases ( most notably cdc48 ) have been shown to result in mitotic arrest [41] . A homolog of the nuclear actin ARP4a is present in the rescue locus of mitotic uncoupling mutant 20C2 . Nuclear actins are involved in transcriptional control and DNA repair and are required for stable attachment of the kinetochore to the mitotic chromosome [42] . Mutations in ARP4a cause defects in the intranuclear spindle and lead to an arrest of the yeast cell cycle in the G2 and mitotic phases . A second set of genes encodes products that harbor protein domains often found in regulatory proteins but which otherwise appear to be unique to Apicomplexa . Gene 583 . m05476 features a TBC domain , an activator domain known to regulate rab-like GTPases and is a motif also found in the yeast spindle factor BUB2p [43] . Two mutants , 42D6 and PO-B3 , were complemented by proteins harboring RCC1-domains ( regulator of chromosome condensation , 25 . m01896 and 72 . m00409 ) . RCC1 proteins control nuclear transport and mitotic progression through nucleotide exchange of ran-GTPases [44] . Interestingly , another unrelated T . gondii RCC1 protein was recently identified as a non-essential protein that when mutated attenuates the virulence of Type I strains in the mouse model [44] . Gene 20 . m03766 , which complements the G1 mutant 109C6 , contains an RNA recognition motif ( RRM ) present in RNA splicing factors , among others . Key features of the mutant 109C6 phenotype share similarities to the defects observed in yeast splicing mutants [45] , and epitope tagged TgRRM-protein exclusively localizes to the tachyzoite nucleus ( not shown ) consistent with a putative role in pre-mRNA splicing . Sequencing of the RRM-protein allele from mutant 109C6 reveals a single base change that alters a highly conserved tyrosine residue in the RRM motif ( 169-Y to N ) , indicating that this mutation is likely responsible for the G1 arrest . In support of this hypothesis , we were unable to complement mutant 109C6 with a gene fragment carrying the mutant 20 . m03766 allele .
Mutant analysis and genetic complementation are powerful strategies to link specific genes to biological pathways [46–48] . Efforts to develop these methods in Toxoplasma have resulted in an episome-based protocol [49] and an insertion based approach [31] . In the latter , phage recombination [50] was used to mobilize cDNA fragments integrated into parasite transformants [30–32] . While these earlier methods achieved some success , the redundancy inherent to cDNA libraries limited the complementation success ( B . Striepen and M . White , unpublished data ) . The genomic-DNA based approach introduced here benefits from the large inserts carried by cosmid vectors to deliver genes in their natural chromosome organization . As a consequence of these improvements , our success rate for genetic rescue of ts growth mutants has improved dramatically . Here we report on the complementation of a diverse selection of mutants ( >20 mutants , with a failure rate <5% ) . The genes identified in these experiments represent a wide range of coding and genomic sizes . Consistent with a mostly regulatory role of their products , the level of transcript for these genes is generally modest based on microarray analysis of tachyzoite gene expression ( see summary in Figure S2 ) . One of the mutants complemented in this study ( 11C9 ) was the subject of an earlier cDNA-based complementation experiment that yielded a suppressor ( TgXPMC2 ) of the genetic defect in this mutant [32] . Multiple complementation experiments of 11C9 using the cosmid library have repeatedly identified a single Toxoplasma gene , 50 . m03077 ( Table 1 ) , which was likely underrepresented in the previously used cDNA libraries due to its large size and low expression level ( <100 units average fluorescence intensity in tachyzoite microarrays for 50 . m03077 versus ∼15 , 000 units for GRA-1 , which was the promoter source used to drive expression in the cDNA libraries [31 , 32] ) . Likewise , we have not re-isolated TgXPMC2 by cosmid complementation . Because gene expression from cosmids relies on native regulatory regions , we believe there is a higher likelihood that the genes identified through this approach will represent the defective gene rather than a suppressor . In the three ts mutants ( 109C6 , VA-15 , and FV-P6 ) where this question was examined so far sequencing and functional testing of the corresponding mutant alleles confirms this prediction . In summary , the cosmid system provides robust complementation for a broad range of genes . Furthermore , the identified loci can be readily biologically validated taking advantage of an extensive set of end-sequenced and tiled cosmids that provide essentially full genome coverage . The protocols and reagents developed in the course of this study should allow future forward genetic analysis of any essential aspect of parasite biology for which a mutant screen can be devised . Tachyzoite growth rates differ dramatically among parasite strains and growth rate is a key virulence determinant in Toxoplasma [22 , 51] . Despite the obvious importance of growth control , how the parasite regulates growth and cell division remains largely unknown . A series of defined biochemical controls and checkpoints regulating progression through one cell cycle phase to the next have been established for a variety of eukaryotic models [52–54] . The catastrophic break down of cell cycle coordination observed in T . gondii in the course of certain drug treatments has lead to the hypothesis that there might be significantly fewer cell cycle controls in this microrganism [29 , 55] . By contrast , the phenotypic groups that have emerged from the collection of conditional growth mutants described in this study support the notion of specific mechanisms and checkpoints . For example , two ts mutants were isolated that reversibly arrest in the G1 phase when shifted to 40°C ( mutant 63H4 and 31F1 ) . The presence of such a natural G1 checkpoint is further supported by the observation that end-stage differentiated parasite forms ( sporozoite and bradyzoite ) show a uniform haploid DNA content [28 , 56] , as do parasites that have been treated with the G1 phase inhibitor pyrrolidine dithiocarbamate [57 , 58] . Tachyzoites released from this drug block , grow synchronously through at least two division cycles , indicating that pyrrolidine dithiocarbamate is likely acting on the same G1 checkpoint affected in our mutants . Another important checkpoint controls entry into S phase , and in Saccharomyces , this checkpoint ( called START ) also controls the initiation of spindle formation and budding [59] . We have previously argued that the tachyzoite cell cycle likely has a similar checkpoint based on the observation that dNTP depletion arrests tachyzoite growth at the G1/S boundary ( 1N DNA content , centrosomes largely duplicated but not yet separated [22 , 28] ) . We have isolated five ts mutants ( e . g . mutant 150B8 ) that display a very similar phenotype at the restrictive temperature , suggesting that the G1/S transition is an important restriction point in the tachyzoite cell cycle as it is in yeast . Mitosis in Apicomplexa has several unique features: the nucleus remains intact , the intranuclear spindle ( s ) reside in a peculiar elaboration of the nuclear envelope the so called centrocone , and daughter cells are scaffolded as internal buds which develop in close proximity and likely under the control of the extranuclear centrosomes [1 , 60] . There is significant evidence for the tight regulation of mitotic events in tachyzoites from three groups of mutants in our collection . These mutants arrest either in mitosis ( 11C9 ) , display defects in chromosome segregation ( 5 mutants ) , or loose the coordination of karyokinesis with cytokinesis at various stages in the replication timeline ( 12 mutants ) . Mitotic mutant V-A15 becomes both aneuploid and polyploid at the restrictive temperature and fails to initiate internal budding . The gene affected in this mutant is an NIMA-related serine/threonine kinases ( Nek ) . This kinase family was first identified as essential for division in Aspergilus nidulans [61] and its members have since been identified as cell cycle regulators throughout eukaryotes [62 , 63] , including protozoa [35 , 36 , 64] . Neks have roles in microtubular dynamics in cilia , mitotic spindles and centrioles [63 , 65 , 66] . Consistent with the known roles for NIMA-related kinases , the mutation in V-A15 leads to defects in the spindle apparatus ( reflected in the loss of MORN1 organization ) and this causes chromosome mis-segregation . Apicomplexa encode a family of related NEK proteins , and in P . falciparum these genes have been shown to be expressed in a developmentally regulated fashion , making it likely that NEKs are critical to fine-tuning the cell cycle to different life-cycle stages and host cells . At the non-permissive temperature mutants 42D6 and PO-B3 are promiscuous for nuclear reduplication leading to the formation of syncytial cells with multiple nuclei . Daughter budding is also abnormal in these mutants and uncoupled from the controls that ensure proper nuclear sorting into each daughter . Two distinct RCC1 domain proteins were found to rescue these mutants ( 25 . m01896 and 72 . m00409 ) . In other eukaryotes , RCC1 domain proteins interact with Ran-GTPases to regulate spindle assembly as well as other mitotic progression controls through modulation of nuclear trafficking [67] . Like mutant V-A15 , mutant PO-B3 ( and also uncoupling mutant 42D6 ) looses MORN1 organization at the restrictive temperature pointing to a potential spindle defect ( data not shown ) . Overall the phenotype of this mutant , as well as several other members of the uncoupling class produced here ( e . g . 42D6 , 20C2 , and 7A11 ) , are similar to the abnormal daughter budding and the induction of unregulated nuclear replication associated with the disruption of the parasite spindle by pharmacological microtubule ablation [29] . Collectively , these observations indicate that proper control over chromosome copy number and budding in Apicomplexa might critically rely on an intact intranuclear spindle or associated structures . In this context it is important to note that the unique centrocone structure that conducts the apicomplexan spindle into the nucleus appears to persist throughout the cell cycle at least in some Apicomplexa [20 , 68] . This model is further supported by preliminary electron microscopy studies of mitotic mutant 11C9 , which upon temperature arrest retains an intact spindle and early daughter scaffolds ( [32] and S . Halonen and M . White , unpublished data ) and does not undergo nuclear reduplication as is seen in mutants 42D6 and PO-B3 . Thus , these models predict that rather than the absence of cell cycle controls in the Apicomplexa , mitotic control mechanisms might require a strict physical context associated with the centrosome and/or the centrocone of the parasite nucleus . Breaking this physical context by drug treatment [29] , overexpression of a centrocone structural componenent [20] , or mutation ( as in the uncoupling group of ts mutants ) results in catastrophic loss of regulation . There is considerable precedence for spatial control of cell cycle checkpoint proteins through compartmental exclusion as well as physical tethering of factors to the centrosomes and spindle structure [69] . Strict compartmentalization of the cell cycle machinery could be a key to the cell cycle flexibility observed in these parasites . Further work is needed to validate this hypothesis . However , the large collection of mutants and genes identified in this screen provides an important pool of validated candidates for mechanistic dissection . Studies that link parasite cell cycle control to the adaptation to specific host cell niches and pathogenesis will be of particular interest .
Parasites were grown in human foreskin fibroblasts ( HFF ) as described [70] . All transgenic and mutant parasite lines are derivatives of the RH parasite line . The non-fluorescent , visually screened mutants were generated from a RH-hxgprt-parasites parent line . Parasites grown overnight at 37°C in 150 mm plates are mutagenized with 400 μg/ml ethyl-nitroso-urea ( ENU ) for 4 h at 37°C ( ∼70% killing ) [9] . The plates were washed to remove the mutagen and the parasites harvested by needle-passage and purified by filtration . Purified parasites are diluted to 40 parasites/ml and 0 . 15 ml/well plated directly into 96-well plates , which yields ∼33 single plaques per plate line ( Poisson distribution predicts 37 . 5% maximum ) . Following 10 days of growth , the master plates are scored for single colonies and then replica plated into two test plates ( 34°C and 40°C ) using a 96-pin array that transfers 20 μl per pin ( VP 403E Grooved Pin replicator , V&P Scientific ) [13] . Growth/no-growth at 34°C and 40°C is scored 7 days later and putative ts mutants are passed into duplicate 24-well plates to reconfirm temperature-sensitive phenotype ( secondary test ) . Each screen of 200–96-well plates yielded ∼6 , 000 single ENU clones ( see Table S1 ) . Cytoplasmic YFP expressing parasite mutants are based on the 2F-1 YFP2 line parent line [13] . In addition , these parasites stably express β-galactosidase . Chemical mutagenesis was performed as described above . Following mutagenesis and release from the host cells , parasites were resuspended in medium without phenol red at a concentration of 250 parasites/ml and cloned directly in 384-well plates ( transparent TC-coated plates; NUNC , Roskilde , Denmark ) confluent with HFF cells using a plate filler at 40 μl/well ( Q-fill: Genetics , New Milton , UK ) . Parasites were expanded for 7 days at 35°C . To suspend parasites , the plates were shaken 1 min at maximum speed on a Bellco orbital shaker ( Vineland , NJ ) placed inside a laminar flow hood . Parasites were grown for six more days until approximately 80% of the monolayer was lysed . Plates were shaken again and duplicated into two identical copies in black-with-optical-bottom TC treated 384-well plates ( Falcon/BD , San Jose , CA ) using a custom-made 384-pin tool with 5 μl slots ( VP Scientific , San Diego , CA ) . Plate copies were preseeded with HFF cells and 40 μl of medium without phenol red . One plate was kept at 35°C and the second copy was incubated at 40°C . After 4 days YFP fluorescence was measured in each well using a FluoStar plate reader ( BMG , Offenburg , Germany ) as described [13] . Differential relative fluorescence units of wells in the 35°C versus the 40°C plate were calculated using a script in Excel ( Microsoft , Redmond , WA ) . All identified ts clones were expanded into T25 flasks at 35°C and their growth arrest at 40°C was confirmed by microscopy in a 24-well plate format . Parasite reversion frequencies were measured by plaque assay at 34°C versus 40°C using several parasite dilutions ( 107-104 parasites/T175cm2 flask ) . DNA content and cellular and nuclear morphologies at the permissive and restrictive temperatures were used to characterize the phenotypes of individual ts-clones . A culture of parasites in log phase ( 8–16 parasites per vacuoles ) was used to seed duplicate T75cm2 flasks ( 5x106 parasites/flask ) for Sytox-Green flow cytometry and to inoculate 6 well plates with coverslips ( 1x106 parasites/well ) for immunofluorescence assays . Following a 2 h incubation at 34°C , the cultures were washed 3x , allowed to grow at 34°C for 5 h and then one set of cultures was shifted to 40°C . Co-staining of infected cultures with a monoclonal mouse antibody 45 . 15 for IMC1 ( 1:2000 , a gift from Dr . Ward , University of Vermont ) and a polyclonal rabbit antiserum for TgPCNA1 ( 1:5000 ) [71] was used to examine changes in cell morphologies and to determine the presence of internal daughter buds and to establish whether replication foci indicative of active DNA replication were present . Where indicated , further antibodies used are: monoclonal 20H5 specific for centrin1 ( 1:2000; a generous gift from Dr . Salisbury ( Mayo Clinic , Rochester , MN ) [72]; rabbit anti-TgMORN1 ( 1:200 ) [20]; rabbit anti-TgIMC3 ( 1:1000 ) [25]; rabbit anti-GFP ( 1:5000; Torrey Pines Biolabs , CA ) ; monoclonal antibody 12G10 anti-α-tubulin ( 1:10; a kind gift from Jacek Gaertig Univ . of Georgia , [73] ) . 4' , 6-diamidino-2-phenylindole ( DAPI ) staining was also performed to judge the qualitative changes in DNA content in parasite clones growth at 34°C versus 40°C . Briefly , parasite cultures grown on coverslips in 6-well plates were fixed with 3 . 7% paraformaldehyde ( pH7 . 4 ) and then permeabilized with 0 . 25% Triton X-100 . Coverslips were washed and incubated in 1X PBS pH 7 . 4 containing 5% FBS and 3% BSA ( blocking solution ) for at least 30 min . Primary antibodies diluted in blocking solution were incubated for 30–60 min , the coverslips washed 3X with blocking buffer and then incubated for 30–60 min with secondary antibodies: Alexa Fluor 594 for mouse and Alexa Fluor 488 for rabbit diluted 1:2000 in blocking buffer . The secondary antibodies were removed and a DAPI solution ( 1 mg/ml stock ) diluted 1:100 in blocking buffer was added for 5–10 min prior to washing and mounting the coverslips in gel mount . Parasites are evaluated with an epifluorescence microscope ( Eclipse TE300 , Nikon Inc . , Melville NY ) and images collected with a digital camera ( SPOTTM , Dynamic Instruments Inc ) . In addition , pictures were acquired on a DM IRB inverted microscope ( Leica ) equipped with a PLAPO 100x/1 . 4 lens and as well as an Axiovert 200M inverted microscope ( Zeiss ) equipped with a PRIOR stage and a NEOFLUAR 100x/1 . 3 lens . The latter two microscopes are equipped with a Hamamatsu C4742–95 CCD camera , and both are controlled by Improvision software ( Lexington , MA ) . Parasite nuclear DNA content was evaluated by flow cytometry using SYTOX Green ( Invitrogen ) staining of tachyzoites grown at the permissive and restrictive temperatures [2] . Briefly , parasites were harvested from the T75cm2 flasks by needle passage and filtration through 3 μm filters . Parasites collected by centrifugation were resuspended in 300 μl cold PBS and 700 μl of cold 100% ethanol added drop-wise . The fixed samples were stored at −20°C for at least 24 h prior to staining for flow cytometry . Fixed parasites are pelleted at 3000 x g , resuspended in 50 μM Tris pH 7 . 5 at a final concentration of 6 × 106 parasites/ml and stained with SYTOX Green ( 1 μM ) . RNase cocktail ( 250 U; RNase A , RNase T1 ) was added and the parasites incubated in the dark at room temperature for 30 min . Nuclear DNA content was measured based on fluorescence ( FL-1 ) using a 488 nm argon laser flow cytometer . Fluorescence was collected in linear mode ( 10 , 000 events ) and the results were quantified using CELLQuestTM v3 . 0 ( Becton-Dickinson Inc . ) . The percentages of each cell cycle phase were calculated based on defined gates for each population . A cosmid library was constructed in ToxoSuperCos . This new cosmid backbone plasmid is composed of the double cos-sites from plasmid SuperCos ( Stratagene , La Jolla , CA ) combined with the kanamycin resistance gene and the origin of replication from pDONR201 ( Invitrogen , Carlsbad , CA ) , the drug pyrimethamine resistant allele of T . gondii dihydrofolate reductates-thymidylate synthase ( DHFR-TSm2m3 ) from pDHFR-TSc3ABP [31] and a synthetic multiple cloning site containing restriction sites for KpnI , HindII , BglII and XhoI . Briefly , pDONR201 was digested with PstI , ends were blunted with T4 DNA Polymerase and digested with ApaI . The resulting 2041 bp fragment was ligated with the DHFR-TSm2m3 expression cassette obtained from pDHFR-TSc3ABP by NotI digestion , T4 DNA polymerase treatment and ApaI digestion as described above . Subsequently , the double cos-sites from SuperCos were introduced in the pDONR-DHFR-TS hybrid by digesting this plasmid with PacI and ApaI and ligation with a PCR product containg the cos-sites ( F-primer ( ApaI ) : ACTGGGGCCCCGTCTTCAAGAATTCGC; R-primer ( PacI ) : ACTGTTAATTAAAGGCTCTCAAGGGCATCGG ) using SuperCos as a template digested with PacI and ApaI . Lastly , a polylinker was introduced into the SpeI site by hybridizing two synthetic oligonucleotides: F-hyb: CTAGTGGTACCAAGCTTAGATCTCTCGAGA and R-hyb: CTAGTCTCGAGTGATCTAAGCTTGGTACCA followed by ligation into the plasmid opened by SpeI restriction . The resulting ToxoSuperCos plasmid is shown in Figure 5 . To construct a cosmid library , genomic DNA was extracted from wild type RH strain . Parasites were lysed in 0 . 2% SDS in 10 mM Tris/HCl pH 8 . 0 followed by digestion with 0 . 8 μg/ml proteinase K and 0 . 04 μg/ml RNaseA for 4 hrs at 37°C . DNA was extracted once with phenol-chloroform-isoamylalchohol ( 25:24:1 ) followed by chloroform extraction and precipitation using 0 . 2 volumes of 10 M NH4OAc and 2 volumes of 100% ethanol . DNA was spooled out using a glass rod and allowed to resuspend in two volumes of TE overnight at 4°C . 5–10 μl of this DNA preparation was digested with a serial dilution of Sau3AI . Reactions were analyzed on 1% agarose gels ( 0 . 5xTBE ) by pulsed field gel electrophoreses on a BioRad CHEF system ( 15 hrs at 4°C using conditions optimized for separation from 5–200 kb at a gradient of 6 . 0 V/cm and an angle of 120° ) using the New England Biolabs ( Beverly , MA ) Low Range PFGE marker as size standard . Conditions generating the highest amount of fragments between 40–55 kb ( as well as one dilution step above and below ) were chosen for preparative Sau3AI digestion . After verification by CHEF , the three digests were pooled and treated as described in the SuperCos manual ( Stratagene ) . Briefly , digested genomic DNA was phenol extracted , treated with calf intestinal phosphatase ( CIP ) and phenol extracted again . The ToxoSuperCos plasmid was prepared by digestion with XbaI , CIP treatment and BamHI digestion . Genomic DNA ( 2 . 5 μg ) and plasmid ( 1 μg ) were ligated overnight and packaged with the Gigapack III XL packaging extract ( Stratagene ) selecting for inserts between 47 and 51 kb . Cosmid-Phages were infected into XL-blue MRF' E . coli and plated on 10 μg/ml LB-kanamycin plates . The resulting library had a complexity of 1 . 25 x106 independent colonies and was frozen as a bacterial stabilate . To produce cosmid library DNA for parasite complementation transfections the library was amplified by plating the stabilate at a density of 104 colonies per 150 mm diameter Petri dishes ( 10 μg/ml LB-Kan; 50 dishes per batch ) , colonies were scraped and cosmid DNA was isolated using the Qiagen ( Hilden , Germany ) Large-Construct Kit according to the manufacturer's instructions . 25 μg cosmid DNA was used for electroporation of 8 x107 parasites prior to inoculation into T175cm2 flasks . Parasites were allowed to recover for 24 hrs at 35°C prior to shifting the to 40°C and addition of 1 μM pyrimethamine to select for phenotypic complementation and cosmid backbone genomic integration , respectively . Individual cosmids from the tiled and sequenced plates were isolated from 5 ml bacterial cultures . Cells were lysed using 250 μl P1 buffer ( Qiagen ) supplemented with 70 μl 10 mg/ml lysozyme solution , followed by addition of Qiagen bufferes P1 and P3 according to the manufacturer's instructions . Following centrifugation , cosmid DNA was precipitated from the supernatant by addition of 0 . 7 volumes of isopropanol . Genomic DNA was extracted from complemented parasite lines ( polyclonal or clonal lines ) and 2 μg of DNA was digested with one of the restriction enzymes in the linker region ( 20 U XhoI , HindIII , BglII or KpnI ) . Digested DNA was purified using the Quiaquick PCR purification kit ( Qiagen ) and an equivalent of 0 . 3 μg digested DNA was self-ligated ( overnight ) in a volume of 20 μl at 16°C . After phenol extraction and ethanol precipitation , 15% of the total reaction ( 1 . 5 μl of 10 μl ) was electroporated into 20 μl DH12S electromax E . coli ( Invitrogen ) and plated on LB-KAN ( 50 μg/ml , DNA can also be purified using a Qiagen spin column instead of phenol extraction ) . DNA minipreparations were isolated for multiple colonies and digested with BglII/NotI . Inserts were sequenced using the T3 promoter primer . | Parasites of the phylum Apicomplexa cause numerous important diseases , including malaria , toxoplasmosis , and cryptosporidiosis . The ability to modify the genome of these parasites by transfection has been the technological key to unlock the biology of parasitic diseases at a molecular level . In this study we further extend the experimental possibilities for the study of apicomplexans by adapting a classic forward genetic approach for Toxoplasma gondii . We have developed protocols and reagents to generate large numbers of mutant parasites , screens to hone in on a subset of mutants of particular interest , and tools to identify the mutated genes that are responsible for the phenotype . Using this new approach , we have genetically dissected the way the parasite divides and multiplies within its host cell . This effort has yielded a series of highly informative mutants along the progression of the apicomplexan cell cycle and more than 20 genes involved in orchestrating parasite cell division . Importantly , this approach should allow unbiased genetic analysis of any part of parasite biology for which a screen can be devised using the Toxoplasma model . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"apicomplexa",
"toxoplasma",
"toxoplasma",
"gondii",
"eukaryotes",
"microbiology"
] | 2008 | Forward Genetic Analysis of the Apicomplexan Cell Division Cycle in Toxoplasma gondii |
Plasmodium vivax can cause severe malaria , however its pathogenesis is poorly understood . In contrast to P . falciparum , circulating vivax parasitemia is low , with minimal apparent sequestration in endothelium-lined microvasculature , and pathogenesis thought unrelated to parasite biomass . However , the relationships between vivax disease-severity and total parasite biomass , endothelial autocrine activation and microvascular dysfunction are unknown . We measured circulating parasitemia and markers of total parasite biomass ( plasma parasite lactate dehydrogenase [pLDH] and PvLDH ) in adults with severe ( n = 9 ) and non-severe ( n = 53 ) vivax malaria , and examined relationships with disease-severity , endothelial activation , and microvascular function . Healthy controls and adults with non-severe and severe falciparum malaria were enrolled for comparison . Median peripheral parasitemia , PvLDH and pLDH were 2 . 4-fold , 3 . 7-fold and 6 . 9-fold higher in severe compared to non-severe vivax malaria ( p = 0 . 02 , p = 0 . 02 and p = 0 . 015 , respectively ) , suggesting that , as in falciparum malaria , peripheral P . vivax parasitemia underestimates total parasite biomass , particularly in severe disease . P . vivax schizonts were under-represented in peripheral blood . Severe vivax malaria was associated with increased angiopoietin-2 and impaired microvascular reactivity . Peripheral vivax parasitemia correlated with endothelial activation ( angiopoietin-2 , von-Willebrand-Factor [VWF] , E-selectin ) , whereas markers of total vivax biomass correlated only with systemic inflammation ( IL-6 , IL-10 ) . Activity of the VWF-cleaving-protease , ADAMTS13 , was deficient in proportion to endothelial activation , IL-6 , thrombocytopenia and vivax disease-severity , and associated with impaired microvascular reactivity in severe disease . Impaired microvascular reactivity correlated with lactate in severe vivax malaria . Findings suggest that tissue accumulation of P . vivax may occur , with the hidden biomass greatest in severe disease and capable of mediating systemic inflammatory pathology . The lack of association between total parasite biomass and endothelial activation is consistent with accumulation in parts of the circulation devoid of endothelium . Endothelial activation , associated with circulating parasites , and systemic inflammation may contribute to pathology in vivax malaria , with microvascular dysfunction likely contributing to impaired tissue perfusion .
While P . falciparum accounts for a majority of severe and fatal malaria cases worldwide , P . vivax is a major cause of morbidity outside of Africa , causing an estimated 70–390 million malaria cases per year [1] . Although previously considered benign , P . vivax is now recognized as capable of causing severe and fatal disease [2] , [3] , [4] , [5] , [6] , [7] , [8] , [9] , [10] . Despite this , little is known about the pathogenesis of severe disease in vivax malaria . In falciparum malaria , severe and fatal disease is characterised by cytoadherence of parasitized red blood cells ( RBCs ) to activated and dysfunctional endothelium , leading to parasite sequestration with microvascular obstruction [11] , [12] , [13] . As a result of sequestration the mature schizont stages of P . falciparum are rarely seen in peripheral blood [14] . Total parasite biomass is underestimated by circulating parasitemia quantitated by microscopy of peripheral blood , and is more accurately quantitated by plasma P . falciparum histidine rich protein-2 ( HRP2 ) . Pathogenesis and disease severity in falciparum malaria are both biomass-related: in contrast to peripheral parasitemia [15] , [16] , [17] , plasma HRP2 is strongly and independently correlated with disease severity and mortality among both children [17] , [18] , [19] , [20] and adults [21] , [22] . Cytoadherence of infected RBCs to endothelial cells is 10-fold less in P . vivax infection than in falciparum malaria [23] , and autopsy evidence for sequestration within the endothelium-lined microvasculature of vital-organs in vivax malaria is minimal [2] , [4] , [24] . This , together with the lower parasitemias resulting from preferential invasion of reticulocytes , is thought to account for the lower lethality of P . vivax [2] . The paucity of apparent endothelial microvascular sequestration in vivax malaria has led to the assumption that peripheral parasitemia reflects total parasite biomass . However , this assumption has been questioned [3] , [25] , [26] , with more adhesive schizont-stages of P . vivax known to be under-represented in peripheral blood [26] , [27] . Accumulation of P . vivax-infected RBCs in the spleen or bone marrow has been hypothesised [3] , [25] , [26] but not yet systematically investigated . In contrast to falciparum malaria , no study has evaluated the relationships between total parasite biomass , disease severity and systemic inflammation in vivax malaria . We propose that in vivax malaria , parasite lactate dehydrogenase ( pLDH ) and P . vivax-pLDH ( PvLDH ) , produced by viable or recently killed parasites , may be used to estimate total parasite biomass . While plasma pLDH has been shown to demonstrate only moderate correlation with peripheral parasitemia in vivax malaria [28] , a pLDH antigen capture enzyme-linked immunosorbent assay demonstrated a direct relationship between parasite production of pLDH and total P . vivax parasite concentration ex vivo , including all parasite stages , suggesting that pLDH reflects total P . vivax parasite biomass [29] . As with HRP2 in falciparum malaria [30] , pLDH is produced to a greater extent by P . vivax schizonts and trophozoites than by ring-form parasites [31] , and given the under-representation of mature P . vivax stages in peripheral blood [26] , [27] , pLDH may be a better marker of total parasite biomass and a better prognostic indicator than peripheral parasitemia . In falciparum malaria , clinical severity and mortality are also independently associated with impaired microvascular function [13] and with increased angiopoietin-2 ( Ang-2 ) [32] , a key product of endothelial Weibel-Palade Bodies ( WPB ) and an autocrine mediator of endothelial activation [33] . Despite the apparent paucity of endothelial microvascular sequestration in vital organs , P . vivax has been associated with greater endothelial activation than P . falciparum , with Ang-2 concentrations higher in non-severe vivax compared to non-severe falciparum malaria [34] . However the relationships between clinical severity , microvascular dysfunction and endothelial WPB exocytosis have not been assessed in vivax malaria . We tested the hypotheses that in vivax malaria , peripheral parasitemia would underestimate total parasite biomass , and that markers of total parasite biomass would be related to systemic inflammation and clinical severity . We also determined the relationship between vivax disease severity and endothelial activation and microvascular function , with patients with non-severe and severe falciparum malaria included for comparison . We found that total vivax biomass was underestimated by peripheral parasitemia , and was associated with both systemic inflammation and disease severity in vivax malaria . Severe vivax malaria was associated with increased endothelial activation and WPB exocytosis and with impaired microvascular function , comparable to that seen in severe falciparum malaria .
A total of 192 malaria patients and 74 healthy controls were enrolled . Malaria patients included 62 with vivax malaria ( 53 non-severe , 9 severe ) and 130 with falciparum malaria ( 109 non-severe , 21 severe ) . The clinical and epidemiological features of 43 patients with vivax malaria ( including 7 with severe disease ) and 122 patients with falciparum malaria ( including 13 with severe disease ) have been previously reported [35] . Baseline characteristics are shown in Table 1 . Among the 9 patients with severe vivax malaria , severity criteria included hypotension ( n = 6 ) , respiratory distress ( n = 1 ) , jaundice ( n = 2 ) , metabolic acidosis ( n = 1 ) , abnormal bleeding ( n = 1 ) and multiple convulsions ( n = 1 ) . Three patients had 2 severity criteria and 6 had one . Pre-antibiotic blood cultures were negative in 4 patients , positive for Streptococcus pneumoniae in one [35] , and not done in 4 . No deaths occurred from either species . In patients with vivax malaria , the median peripheral parasitemia was 2 . 4-fold higher in severe compared to non-severe disease ( 10 , 243 parasites/µL vs . 4 , 209 parasites/µL; P = 0 . 021 ) , while the median PvLDH was 3 . 7-fold higher ( 96 . 6 ng/ml vs . 26 . 2 ng/ml; P = 0 . 021 ) and pLDH 6 . 9-fold higher ( 308 ng/ml vs . 44 . 6 ng/ml; P = 0 . 015 ) ( Table 2 and Fig . 1 ) . After removing the patient with severe vivax malaria and concurrent bacteremia from the analysis , the median peripheral parasitemia was 1 . 8-fold higher in severe compared to non-severe disease ( 7 , 865 parasites/µL vs . 4 , 209 parasites/µL; P = 0 . 047 ) , while the median PvLDH was 4 . 8-fold higher ( 125 ng/ml vs . 26 . 2 ng/ml; P = 0 . 002 ) and pLDH 7 . 3-fold higher ( 326 ng/ml vs . 44 . 6 ng/ml; P = 0 . 002 ) . In falciparum malaria , median peripheral parasitemia and plasma HRP2 were 4 . 2-fold and 7 . 4-fold higher , respectively , in severe compared to non-severe disease ( Table 2 ) . To estimate whether the “hidden parasite biomass” was larger in severe compared to non-severe malaria , we calculated the ratio of plasma pLDH and plasma PvLDH concentration to peripheral parasite density in vivax malaria , and the ratio of plasma HRP2 concentration to peripheral parasite density in falciparum malaria . Among patients with vivax malaria the ratios of PvLDH and pLDH to parasite density ( log ) were 3 . 7-fold and 5 . 7-fold higher , respectively , in severe compared to non-severe disease . The corresponding fold-increase in the ratio of plasma HRP2 to parasite count ( log ) in patients with severe compared to non-severe falciparum malaria was 5 . 4 . Among patients with vivax malaria pLDH and PvLDH were strongly correlated ( ρ = 0 . 90 , P<0 . 0001 ) . However there was only a modest correlation between PvLDH and parasitemia ( ρ = 0 . 27 , P = 0 . 033 ) , and between pLDH and parasitemia ( ρ = 0 . 28 , P = 0 . 028 ) , comparable to the correlation between parasitemia and plasma HRP2 among patients with falciparum malaria ( ρ = 0 . 41 , P<0 . 0001 ) ( Fig . 2 ) . Among patients with vivax malaria who had detectable pLDH and PvLDH levels on enrolment and in whom longitudinal measurements could be performed ( until day 3 or until undetectable ) , levels were undetectable by day 3 in 16/25 ( 64% ) and 22/31 ( 71% ) patients respectively . In ex vivo assay conditions , P . vivax parasites exist as mature schizonts for an estimated 5% of their 48-hour life-cycle [36] . Despite this , only 14/59 ( 24% ) patients with vivax malaria had any circulating schizonts detectable on peripheral blood film , including 2/9 ( 22% ) with severe and 12/50 ( 24% ) with non-severe disease . Of those with schizonts detectable , schizonts comprised <5% of circulating parasites in 11/14 ( 78% ) patients , including one with severe and 10 with non-severe disease . There was a wide variation in the proportion of peripheral P . vivax parasites at ring ( median 33% , range 0–100% , IQR 9–77% ) and trophozoite stage ( median 67% , range 0–100% , IQR 20–91% ) . Among patients with severe and non-severe vivax malaria , there were no differences between the median percentage of rings or trophozoites in peripheral blood . Median concentrations of Ang-2 and von Willebrand Factor ( vWF ) , the two major products of endothelial WPB release , were both increased in patients with severe and non-severe vivax malaria compared to controls ( P<0 . 01 for all comparisons ) , with Ang-2 nearly twice as high in severe compared to non-severe disease ( P<0 . 0001 ) ( Table 3 ) . The median concentrations of the endothelial adhesion receptors E-selectin and ICAM1 were also increased in severe and non-severe vivax malaria compared to controls ( P<0 . 001 for all comparisons ) . In patients with vivax malaria there was no association between age and any of the endothelial activation markers measured . Median concentrations of Ang-2 , E-selectin and ICAM1 were at least as high in severe and non-severe vivax malaria as they were in severe and non-severe falciparum malaria ( Table 3 ) . The median concentrations of the vWF-cleaving protease , ADAMTS13 , and its activity , were lower among patients with severe and non-severe vivax malaria compared to controls ( P<0 . 001 for all comparisons ) , with median ADAMTS13 activity also lower in severe compared to non-severe vivax malaria ( P = 0 . 027 ) . Median concentrations of IL-6 and IL-10 were higher in severe and non-severe vivax malaria compared to controls ( Table 3 ) . In patients with vivax malaria , peripheral parasitemia correlated with the endothelial WPB products , VWF ( ρ = 0 . 53 , P<0 . 0001 ) and Ang-2 ( ρ = 0 . 36 , P = 0 . 004 ) , as well as the endothelial adhesion receptor E-selectin ( ρ = 0 . 33 , P = 0 . 009 ) , and was inversely correlated with activity of the VWF-cleaving protease ADAMTS13 ( ρ = −0 . 31 , P = 0 . 055 ) ( Table 4 ) . Importantly , and in contrast to peripheral parasitemia , no correlation was seen between either of the total parasite biomass markers PvLDH or pLDH , and the endothelial products VWF , Ang-2 , or E-selectin ( Table 4 ) . While not associated with markers of endothelial activation , PvLDH was correlated with the systemic inflammatory markers IL-6 ( ρ = 0 . 27 , P = 0 . 045 ) and IL-10 ( ρ = 0 . 28 , P = 0 . 035 ) , and , in contrast to peripheral parasitemia , was inversely correlated with baseline platelet count ( ρ = −0 . 26 , P = 0 . 045 ) and with platelet nadir ( ρ = −0 . 27 , P = 0 . 035 ) . Among patients with falciparum malaria , biomarkers of systemic inflammation correlated with both peripheral parasitemia and the total parasite biomass marker , plasma HRP2 ( Table 4 ) . In addition , and in contrast to vivax malaria , endothelial activation correlated not only with peripheral parasitemia but also with total parasite biomass ( Table 4 ) . Baseline platelet count and platelet nadir also correlated with both peripheral parasitemia and HRP2 ( Table 4 ) . In patients with vivax malaria , Ang-2 , in addition to correlating with parasitemia , correlated with the adhesion receptors E-selectin ( ρ = 0 . 40 , P = 0 . 001 ) and ICAM-1 ( ρ = 0 . 54 , P<0 . 0001 ) , and inversely with ADAMTS13 antigen ( ρ = −0 . 41 , P = 0 . 010 ) and activity levels ( ρ = −0 . 60 , P = 0 . 0001 ) , with all associations remaining significant after adjusting for parasitemia . ADAMTS13 antigen and activity levels correlated with platelet count ( ρ = 0 . 36 , P = 0 . 025; and ρ = 0 . 56 , P = 0 . 0003 , respectively ) , with this association remaining significant after adjusting for parasitemia . IL-6 , elevated in vivax malaria , is known to be a major inhibitor of ADAMTS13 activity [37] . In vivax malaria , IL-6 was inversely correlated with ADAMTS13 activity ( ρ = −0 . 40 , P = 0 . 013 ) and platelet nadir ( r = −0 . 44 , p = 0 . 0009 ) . No association occurred between lactate and ADMATS13 antigen and activity levels , or between lactate and Ang-2 . Microvascular reactivity ( assessed using Near Infrared Resonance Spectroscopy [13] ) was decreased in patients with severe vivax malaria ( median 5 . 44 units/sec ) compared to those with non-severe vivax malaria ( median 6 . 98 units/sec; P = 0 . 006 ) and to controls ( median 6 . 34 , P = 0 . 027 ) ( Table 3 and Fig . 3 ) . Microvascular reactivity was also impaired among patients with severe falciparum malaria ( median 4 . 88 units/sec ) compared to those with non-severe falciparum malaria ( 6 . 31 units/sec; P = 0 . 024 ) and controls ( P = 0 . 016 ) , with the degree of impairment similar between severe vivax and severe falciparum malaria . In both species , there was no significant difference between controls and patients with non-severe malaria . Among patients with vivax malaria , microvascular reactivity was inversely associated with the perfusion marker , lactate , in severe ( ρ = −0 . 74 , P = 0 . 04 ) , but not non-severe disease ( ρ = 0 . 09 , P = 0 . 62 ) . In severe disease ADAMTS13 antigen and activity levels correlated with microvascular reactivity ( ρ = 0 . 64 and P = 0 . 09 , for both associations ) , although no association occurred in non-severe disease ( ρ = 0 . 07 , P = 0 . 78; and ρ = −0 . 02 , P = 0 . 95; respectively ) . No association was found between microvascular reactivity and other variables in vivax malaria , including parasitemia or biomass markers . In patients with falciparum malaria , microvascular reactivity was inversely correlated with parasitemia ( ρ = −0 . 22 , P = 0 . 035 ) , and positively correlated with platelet count ( ρ = 0 . 017 , P = 0 . 017 ) . Among patients with severe and non-severe falciparum malaria , no association occurred between microvascular reactivity and lactate , or between microvascular reactivity and ADAMTS13 antigen or activity . In both species there was no association between microvascular reactivity and age . Among patients with vivax malaria , 6 patients with severe and 16 with non-severe disease had NIRS repeated on day 3 , with no change from baseline noted in either group ( median microvascular reactivity on day 3 = 5 . 59 units/sec and 7 . 39 units/sec respectively ) . In severe falciparum , microvascular reactivity improved from 4 . 88 units/sec at enrolment to 6 . 51 units/sec on day 3 ( P = 0 . 039 ) , although no difference was noted among 37 patients with non-severe falciparum malaria ( median microvascular reactivity on day 3 = 6 . 56 units/sec , P = 0 . 437 )
Peripheral parasitemia underestimates measures of total parasite biomass in vivax malaria and circulating schizonts are under-represented in peripheral blood , suggesting tissue accumulation of a hidden biomass of P . vivax-infected RBCs . The proportions of total parasite biomass detectable in peripheral blood in uncomplicated and severe disease were of comparable magnitude in vivax and in falciparum malaria . The association between P . vivax peripheral parasitemia and total parasite biomass is modest , with total parasite biomass more strongly associated with disease severity compared to peripheral parasitemia . The smaller proportion of total P . vivax biomass detectable in peripheral blood in severe compared to non-severe disease suggests that hidden biomass is greater in severe vivax malaria and contributes to pathogenesis of disease . The association of total vivax parasite biomass with systemic inflammation suggests that the hidden biomass is capable of mediating systemic inflammatory pathology . In contrast to peripheral P . vivax parasitemia , total parasite biomass did not correlate with endothelial activation , and we therefore speculate that accumulation of P . vivax-infected RBCs may occur in parts of the circulation devoid of endothelium . In falciparum malaria , peripheral parasitemia underestimates the total parasite biomass due to sequestration of parasitized RBCs within endothelium-lined microvasculature . While P . vivax is not thought to sequester in the endothelium-lined microvasculature to the same degree as P . falciparum , limited histopathological reports show intact P . vivax-infected RBCs in the bone marrow [38] , [39] , [40] , [41] , [42] , [43] and spleen [44] , [45] , organs containing circulatory compartments that are not endothelium-lined . The spleen is a lymphoid organ whose primary role is to clear abnormal erythrocytes from the circulation , and hence plays a fundamental role in removing parasitized RBCs , especially in falciparum malaria where RBC deformability is markedly decreased . In falciparum malaria , endothelial cytoadherence of infected RBCs enable the parasite to sequester in non-splenic microvasculature and avoid passage through the spleen . In contrast , P . vivax-infected RBCs cytoadhere with less avidity to endothelial cells [23] , and demonstrate increased deformability [46] , [47] , allowing circulation through the spleen . While 80-90% of splenic blood flow occurs through endothelial-lined sinuses , the remainder circulates through a parallel slow , open circulation in the splenic cords , devoid of endothelial cells [48] . In a recent report involving a splenectomised patient with vivax malaria , large numbers of intact non-phagocytosed P . vivax-infected reticulocytes were found in the splenic cords [44] . Accumulation of P . vivax-infected reticulocytes in the slow open microcirculation of the spleen , possibly through reticulocyte adherence to non-endothelial resident cells [49] , [50] , [51] , [52] , has been proposed as a mechanism by which P . vivax may avoid macrophage clearance [53] yet readily invade new reticulocytes . The findings of our study suggest that tissue accumulation of P . vivax does occur and is greatest in severe vivax malaria , and we speculate it may occur in a location devoid of endothelium , such as the slow circulation of the spleen . We found no correlation between the P . vivax biomarkers of total biomass , pLDH and PvLDH , and either of the two products of endothelial WPB release , Ang-2 or VWF , or the endothelial adhesion receptor E-selectin . In contrast , peripheral parasitemia correlated with endothelial activation ( Ang-2 , VWF and E-selectin ) in both vivax and falciparum malaria , and with total biomass ( HRP2 ) in falciparum malaria . These findings suggest that endothelial cells may be activated by P . vivax-infected RBCs circulating through the peripheral microvasculature , but not by the hidden P . vivax biomass . Accumulation of P . vivax in the endothelium-free open circulation of the spleen may account for these findings . The finding that P . vivax peripheral parasitemia and markers of total parasite biomass both correlated with the leukocyte-derived cytokines may reflect the ability of circulating and tissue-accumulated P . vivax to stimulate leukocytes in both peripheral blood and leukocyte-containing organs such as the spleen or bone marrow , and thereby mediating organ dysfunction secondary to systemic inflammation . Although underestimating total biomass , peripheral parasitemia was still higher in severe compared to non-severe malaria in both P . vivax and P . falciparum , and endothelial activation was higher in patients with severe disease . Ang-2 concentrations were twice as high in severe compared to non-severe vivax malaria , and comparable to levels seen in severe falciparum malaria . Ang-2 , which causes autocrine endothelial activation , has previously been shown to be markedly elevated in severe falciparum malaria and a consistent predictor of death in both adults [32] , [54] , [55] and children [56] , [57] . In severe falciparum malaria Ang-2 correlates with impaired endothelial function , lactate , plasma HRP2 , ICAM-1 , and E-selectin [32] . In our study , Ang-2 was also associated with ICAM-1 and E-selectin among patients with vivax malaria , consistent with the role of Ang-2 in activating endothelium in vivax as well as falciparum malaria [32] . Our findings of increased endothelial activation in severe vivax malaria are consistent with a recent study reporting higher concentrations of ICAM-1 and VCAM-1 in severe compared to uncomplicated vivax malaria [58] . Microvascular function was significantly impaired in severe vivax malaria , comparable to the impairment seen in severe falciparum malaria . As previously shown in severe falciparum malaria [13] , the impairment of microvascular reactivity in severe vivax malaria was strongly associated with blood lactate , suggesting that impaired tissue perfusion is contributing to organ dysfunction in severe vivax malaria . Plasma ADAMTS13 activity was decreased in vivax malaria , with deficiency associated with both severe disease and with impaired microvascular function . ADAMTS13 is a protease that cleaves ultra-large and prothrombogenic VWF multimers ( UL-VWF ) , and deficiency leads to accumulation of UL-VWF with resultant increase in platelet aggregation and adhesion , and microvascular thrombosis . Accumulation of UL-VWF resulting from ADAMTS13 deficiency is characteristic of the microangiopathic disease thrombotic thrombocytopenic purpura ( TTP ) , and has been reported in patients with both falciparum and vivax malaria [59] . While renal failure was not a feature of patients with severe vivax malaria in our study , thrombotic microangiopathy has been reported in severe vivax disease elsewhere [60] , [61] . In our study , lower ADAMTS13 antigen and activity levels were associated with lower platelet counts , and with increased concentrations of IL-6 , a known specific inhibitor of ADAMTS13 activity [37] . We speculate that biomass-related IL-6 may contribute to impaired ADAMTS13 activity and accumulation of UL-VWF multimers , and may thereby contribute to microvascular dysfunction , thrombocytopenia and disease severity in vivax malaria . Further studies however are required to confirm the mechanisms underlying ADAMTS13 deficiency in vivax malaria , and to investigate the role of thrombotic microangiopathy . Our study had several limitations . Firstly , pLDH and PvLDH have not been previously validated as biomass markers in vivax malaria . Other factors may contribute to the plasma concentrations of these markers , such as host metabolic clearance rate , distribution within the host and diffusion rates from host tissue , and natural variation in parasite expression . Release of pLDH solely at the time of schizont rupture may provide a possible alternative explanation for the association between concentrations of pLDH/PvLDH , systemic inflammation and disease severity , with schizogony also associated with an inflammatory response [62] . However , we have demonstrated that ex-vivo concentrations of PvLDH increase progressively during short term culture of P . vivax , indicating that PvLDH is secreted throughout the parasite life-cycle and supporting the use of pLDHand PvLDH as surrogate markers of parasite biomass . Host inflammatory response may contribute in part to the marked under-representation of P . vivax schizonts in peripheral blood , however this would apply similarly to hosts infected with other Plasmodium species with under-representation of schizonts in peripheral blood , such as P . falciparum or P . coatneyi , whose schizonts have been clearly shown to sequester in tissues [63] . The small number of patients with severe vivax malaria ( n = 9 ) limited our ability to evaluate the association between disease severity and other variables , and to assess for associations of parasite biomass and/or parasitemia that may have occurred only among patients with severe disease . In addition , with multiple comparisons we are unable to exclude the possibility that associations may have occurred by chance . However , the magnitude and direction of the associations with disease severity and with different measures of biomass are consistent , plausible , and are comparable to those in patients with falciparum malaria . Pre-antibiotic blood cultures were not performed in all our patients , and hence it is possible that as in falciparum malaria [64] , [65] , concurrent bacteremia may have contributed to the clinical features of some patients with severe vivax malaria . However , the much higher biomass in patients with severe compared to non-severe vivax cannot have been accounted for by concurrent bacteremia , or the parasitemia- or biomass-related associations we found . Finally , while we demonstrated increased endothelial activation among patients with vivax malaria in proportion to disease severity [34] , the causes of WPB exocytosis in vivax malaria were not evaluated and require further study . In summary , peripheral parasitemia underestimates total parasite biomass in vivax malaria , particularly in severe disease , suggesting tissue accumulation of a hidden biomass of P . vivax-infected RBCs . While histological studies are needed to confirm this , we propose that a hidden P . vivax biomass is capable of mediating systemic inflammation which may contribute to organ dysfunction . The correlation of markers of total P . vivax biomass with systemic inflammation but not with markers of endothelial activation is consistent with the hypothesis that accumulation of P . vivax-infected red cells may occur in parts of the circulation devoid of endothelium , such as the open circulation of the spleen or extravascular bone marrow . In contrast only peripheral parasitemia was associated with endothelial activation and the Weibel Palade Body products Ang-2 and VWF . The association between clinical severity and endothelial activation , ADAMTS13 deficiency , thrombocytopenia , and impaired microvascular function suggests that thrombotic microangiopathy , systemic inflammation and microvascular dysfunction may contribute to pathogenesis of disease in vivax malaria .
This study was conducted in Sabah , Malaysia , a region currently in the pre-elimination phase of malaria control and where P . vivax endemicity is low [66] , [67] . Patients were enrolled as part of a prospective clinical and epidemiological study of all malaria patients admitted to Queen Elizabeth Hospital , an adult tertiary referral hospital [35] . Consecutive patients with PCR-confirmed vivax or falciparum monoinfection were enrolled from September 2010 – October 2012 ( with non-severe falciparum malaria patients included up until October 2011 ) if they were non-pregnant , ≥12 years old , had no major comorbidities or concurrent illness , were within 18 hours of commencing antimalarial treatment , had haemoglobin >7 . 0 g/dL , and had not been previously enrolled in the study . Clinical details of patients enrolled from September 2010 – October 2011 , in addition to details regarding excluded patients , have been previously reported [35] . Severe malaria was defined as the presence of ≥1 of: unrousable coma ( Glasgow Coma Scale score <11 ) ; multiple ( >2 ) convulsions; respiratory distress ( respiratory rate >30 breaths per minute and oxygen saturation <94% ) ; hypotension ( systolic blood pressure ≤80 mm Hg ) ; jaundice ( bilirubin >43 µmol/L plus parasitemia >20 000/µL [P . vivax] or >100 , 000 [P . falciparum] and/or creatinine >132 µmol/L ) ; significant abnormal bleeding; hypoglycaemia ( blood glucose <2 . 2 mmol/L ) ; metabolic acidosis ( bicarbonate <15 mmol/L or lactate >4 mmol/L ) ; acute kidney injury ( AKI; creatinine >265 µmol/L ) ; hyperparasitemia ( P . falciparum parasitemia >10% ) . Healthy controls were visitors or relatives of malaria patients , with no history of fever in the past 48 hours and with blood film negative for malaria parasites . Standardized history and physical examination were documented . Haematology , biochemistry , acid-base parameters , and lactate ( by bedside blood analysis; iSTAT system ) were obtained on admission . Parasite counts were determined by microscopy , and parasite species were identified by PCR [68] , [69] . Patients were treated according to hospital guidelines , as previously described [35] . Venous blood collection in lithium heparin was centrifuged within 30 minutes of collection and plasma was stored at −70°C . Plasma concentrations of the endothelial activation markers ICAM-1 , E-selectin , and Ang-2 were measured using ELISA ( R&D Systems ) , and IL-6 and IL-10 were measured by flow cytometry ( BD Cytometric Bead Array ) . Antigen concentrations of vWF and the vWF-cleaving enzyme , ADAMTS13 , were measured in platelet poor plasma using ELISA ( America Diagnostica ) , and ADAMTS 13 activity by fluorescence resonance energy transfer ( FRET ) , as previously described [70] . Peripheral blood parasitemia was quantitated per 200 white blood cells by microscopy and expressed as parasites/µL based on automated white cell count . Parasite stage distribution was quantitated on microscopy of peripheral blood . Plasma HRP2 ( for P . falciparum ) [32] , genus-specific pLDH ( for P . falciparum and P . vivax ) and PvLDH ( for P . vivax ) were measured by ELISA [28] as proxies for total parasite biomass . Ratios of plasma pLDH and PvLDH to peripheral parasite density were expressed as ngs/1000 peripheral parasites , with the denominator log-transformed . To confirm evidence of PvLDH secretion across the parasite life-cycle , four cryopreserved P . vivax isolates with a high proportion of ring-stages were thawed and cultured ex-vivo as previously described [71] , over 48–54 hours . Starting parasitemias were 3623/µL ( 98% rings ) , 10 , 605/µL ( 98% rings ) , 19 , 342/µL ( 65% rings ) , and 24 , 680/µL ( 98% rings ) . Thick and thin films were prepared at serial time points for stage differential and culture supernatant sampled for concentration of PvLDH . A progressive increase in concentration of PvLDH in culture supernatant across the parasite life-cycle was demonstrated ( Text S1 ) , indicating secretion of PvLDH by all parasite stages . In vitro parasitemia remained stable throughout the culture duration . Microvascular function was assessed using Near Infrared Spectroscopy ( InSpectra 650 , Hutchinson Technology , Hutchinson , MN ) which uses a probe applied to the thenar eminence to noninvasively measure microcirculatory oxygenation ( tissue oxygen saturation; StO2 ) before and after an ischaemic stress , as previously described [13] . To induce an ischaemic stress , a vascular cuff was inflated to 200 mm Hg for 5 minutes and then rapidly deflated . Microvascular function was defined as the gradient of the StO2 recovery slope from release of the vascular cuff until StO2 had reached 85% of the baseline value . Statistical analysis was performed with STATA software ( version 10 . 1; Statacorp , College Station , TX , USA ) . For continuous variables intergroup differences were initially compared using analysis of variance or Kruskal-Wallis tests depending on distribution . Student's T-test or Mann-Whitney tests were used for post-hoc pair-wise comparisons . Categorical variables were compared using χ2 or Fisher's exact test . Associations between continuous variables were assessed using Spearman's ( ρ ) or Peason's ( r ) correlation coefficients , depending on distribution . The study was approved by the Ethics Committees of the Malaysian Ministry of Health and Menzies School of Health Research . Informed written consent was provided by all participating adults , and by the parent or guardian of any participant aged <18 years . | How vivax parasites cause severe malaria is not known . In contrast to falciparum parasites , the number of vivax parasites circulating in peripheral blood is low , and there is thought to be little sequestration of parasitized red cells within endothelium-lined small blood vessels in vital organs . Total parasite burden ( circulating plus hidden ) and activation and dysfunction of the endothelial cells lining blood vessels all contribute to severe disease in falciparum malaria , but have not been evaluated in severe vivax malaria . We measured parasite lactate dehydrogenase ( pLDH ) and P . vivax-pLDH ( PvLDH ) as proxies of total parasite biomass and found that , as in falciparum malaria , the total biomass of vivax parasites is underestimated by counting parasites circulating in peripheral blood , suggesting a hidden burden of vivax parasites . Markers of total vivax biomass were strongly associated with illness-severity and inflammatory cytokines , suggesting that this hidden burden is capable of contributing to generalised inflammation and hence severe disease . Number of peripheral vivax parasites , but not total biomass , correlated with activation of endothelial cells , suggesting that the hidden vivax-infected red cells may accumulate in parts of organs without endothelium , such as the slow-circulation of the spleen or non-blood-vessel parts of the bone marrow . Severe vivax malaria was associated with increased endothelial activation and impaired microvascular function , suggesting that these processes also contribute to impaired blood flow and disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"plasmodium",
"vivax",
"protozoans",
"malarial",
"parasites",
"biology",
"and",
"life",
"sciences",
"malaria",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"organisms"
] | 2015 | Parasite Biomass-Related Inflammation, Endothelial Activation, Microvascular Dysfunction and Disease Severity in Vivax Malaria |
The recent development of novel Polymerase Chain Reaction ( PCR ) technologies that confer theoretical advantages over quantitative PCR has considerable potential in the diagnosis of low load infections , such as Trypanosoma cruzi in the chronic phase of Chagas disease . We evaluated the utility of the digital droplet ( dd ) PCR platform in the detection of T . cruzi infection . We imported a validated qPCR assay targeting the T . cruzi satellite tandem repeat ( TcSTR ) region to the ddPCR platform . Following optimization , we tested and repeated a standard curve of TcI epimastigotes to characterise the analytical performance of the assay on the ddPCR platform . We compared this to published qPCR performance data , and the performance of the qPCR assay in our own testing . We subsequently tested a panel of 192 previously characterized DNA specimens , extracted from the blood of individuals with and without T . cruzi infection . The assay performed well on the ddPCR platform , showing a limit of detection of 5 copies/μL or 1 parasite/mL . This was higher than the published limit of detection for qPCR , which was 0 . 46 parasites/mL . The ddPCR platform was not significantly more accurate than qPCR at any concentration tested . However , the clinical sensitivity and specificity of the assay were both 100% with perfect agreement between qPCR and ddPCR positive and negative result calling in clinical specimens . An average of 9 , 286 copies of TcSTR were detected per parasite . The use of the ddPCR platform to run this assay was comparable , but not superior in terms of performance , to the qPCR platform .
Chagas disease , caused by Trypanosoma cruzi , is a complex chronic pathology that affects around 8 million people worldwide and represents a serious public health problem [1] . T . cruzi parasites exhibit tremendous within-species genetic diversity and have been subdivided by international consensus into at least six discrete typing units ( DTUs ) from TcI to TcVI [2 , 3] . The acute phase of the disease is usually characterized by mild fever , but in a small proportion of cases , it can be accompanied by myocarditis and other lethal complications [4] . Most patients continue into the chronic phase , which can be asymptomatic , but about 30% of the infected patients will develop heart or digestive complications after several years [5] . The diagnosis of Chagas disease is challenging . In the acute phase of infection and disease , the parasitic load is sufficiently high that direct methods of parasitological observation such as a blood smear and hematocrit are recommended to diagnose the disease [6–8] . By contrast , the chronic phase is characterized by a low and intermittent parasitic load that is not likely to be detected using direct parasitological methods . Therefore , serological assays such as indirect immunofluorescence assays , indirect hemagglutination test and enzyme linked immunosorbent assays ( ELISA ) are recommended to diagnose the disease [9–12] . Molecular methods have been developed and currently , there are several real-time quantitative PCR assays that are commonly used internationally to detect and quantify T . cruzi DNA in blood samples of acute and chronic Chagas disease patients [13–15] . The evaluation of a number of different genetic sequences suggested a tandem repeat satellite DNA ( TcSTR ) sequence of 188 bp that is continuously repeated across the genome of T . cruzi to be an appropriate qPCR target for optimal sensitivity and specificity [14] . The operating characteristics of molecular tests for detecting T . cruzi DNA have varied according to clinical phase and technical specifications . The sensitivity for identifying chronic infection of end-point PCR has ranged between 22% and 75% [9 , 16] and in both cases specificity has estimated at 100% . By contrast , for quantitative PCR ( qPCR ) , sensitivity has ranged between 60% and 80% for the chronic phase and between 88% and 100% for the acute phase of infection [6 , 14] . Recently , a loop-mediated isothermal amplification assay ( LAMP ) was reported that was able to detect T . cruzi DNA in peripheral blood samples collected from well-characterised seropositive patients , including acute , congenital , chronic and reactivated Chagas disease [17 , 18] . The sensitivity of qPCR in the acute phase makes it ideal for the detection of such infections; however , the low levels or absence of circulating parasites in chronic cases means that the sensitivity of this assay is currently too low to allow for proactive screening for patients with chronic infection . The development of an assay with better performance at low concentrations may therefore enable the improved detection of infection in chronic Chagas disease cases , which may support case findings and monitoring . Digital droplet PCR ( ddPCR ) is a next-generation PCR technology that enables the absolute quantitation of nucleic acids by separating a reaction into ~20 , 000 compartments ( droplets ) and using the proportion of fluorescent droplets to indicate target occupancy and positivity . By isolating the PCR reaction into compartments , the relative concentration of targets , primers and probes is high , and of inhibitors is low , thereby theoretically allowing improved precision at lower concentrations and an improved analytical limit of detection compared with qPCR . Additionally , ddPCR measures the response to the number of copies present in a linear manner thereby allowing smaller differences in load to be accurately quantified , compared with qPCR that measures logarithmic increases in the target number . The theoretical benefits of this technology have encouraged its evaluation in the detection of pathogens [19] . While ddPCR has been evaluated for the detection of bacteria and viruses ( for example , Chlamydia trachomatis , Salmonella , Escherichia coli , Campylobacter jejuni , cytomegalovirus and HIV ) [20–24] , few investigations have been conducted into the role of ddPCR in the detection of protozoan infections . ddPCR assays have been initiated for Schistosoma , Cryptosporidium and Plasmodium [25–27] . Therefore , the aim of this work was to import a validated international consensus qPCR assay targeting the TcSTR [14] to the ddPCR platform and compare its analytical and diagnostic performance with the qPCR iteration to determine whether the technology may be useful .
The study protocol was approved by the Technical Research Committee and Ethics Research Board at the National Health Institute in Bogotá , Colombia , protocol CTIN-023-17 in the framework of the project “Fortalecimiento y vigilancia de la capacidad diagnóstica de enfermedades emergentes y reemergentes en Colombia” . Participation ( adults only ) was voluntary and patients were asked for informed written consent . The diagnostic evaluation protocol was additionally approved by the London School of Hygiene & Tropical Medicine Observational Ethics Committee ( protocol number 15515 ) . T . cruzi parasites of TcI ( MHOM/CO/04/MG ) and TcII ( MHOM/BR/53/Y ) from different life-cycle stages were grown ( epimastigotes and metacyclic trypomastigotes ) in culture as reported elsewhere [28] . Cultures were aliquoted in the exponential growth stage . A Neubauer chamber was used to assess the concentration of parasites per mL . Then , 300 μL of the counted parasites were subjected to DNA extraction using the High Pure PCR Template Roche kit as reported elsewhere [29] , and the DNA was eluted in 100 μL of H2O or TE buffer . The satellite tandem repeat of T . cruzi ( TcSTR ) was selected due to its high copy number in the genome and its routine usage as a diagnostic target for T . cruzi infection . The TcSTR region can be repeated up to 105 times in the genome and can account for as much as 10% of the genetic material , although this varies between DTUs . The qPCR described by Ramirez et al . was considered an appropriate comparator for this study due to the availability of detailed evaluation data of analytical and diagnostic performance [14] . For this study , qPCR was conducted using published methodology . The standard curve for TcI epimastigote parasite material was determined with the qPCR assay on a single plate using five technical replicates . The TcSTR sequences from Ramirez et al . [14] were used in this study . An endogenous control assay ( Homo sapiens ribonuclease P/MRP subunit p30 [HsRPP30] as described by Luo et al . 2005 [30] ) was also added to the test to enable users to check that specimen collection and sample processing had been completed successfully ( Table 1 ) . Both TcSTR and HsRPP30 assays were run in duplicate in a single well . The ddPCR reaction mixtures ( 20 μL volume ) contained final concentrations of 1× ddPCR Supermix ( Bio-Rad , Hemel Hempstead , UK ) , 0 . 9 μM for each primer and 0 . 2 μM for probes , and 5 μL of purified sample DNA ( Table 1 ) . Droplet generation and droplet reading for ddPCR were carried out using the Bio-Rad QX100 workflow according to the manufacturer's instructions . The first thermal cycling profile tested was matched to the published qPCR protocol ( 95°C for 10 minutes followed by 45 cycles of 95°C for 15 seconds and 60°C for 30 seconds ) . However , under these cycling conditions for ddPCR , the primer–probe set yielded poor separation between positive and negative droplet populations . Based on the results of a systematic assessment of parameters to improve ddPCR performance [31] , the dissociation and annealing/extension steps in the thermocycling conditions were extended to 30 seconds and 2 minutes , respectively , and the number of cycles was increased to 50 . Raw ddPCR data were collected using Quantalife software ( Bio-Rad ) and were then exported for analysis using R scripts as reported elsewhere [20] . Thresholds between positive and negative droplet populations were set manually using per-plate positive and no-template controls as a guide; for the TcSTR assay , droplets with an amplitude above 3 , 000 were considered positive . The Poisson calculation was used to estimate the number of copies/μL of the reaction and confidence intervals [32] . For classification purposes , we followed the method described by Roberts et al . to optimize sensitivity and specificity [20] . We used the estimated mean concentration of the target and its standard deviation to define the cumulative distribution function ( c . d . f . ) at x = 0 . This value describes the probability that the true concentration is less than or equal to zero copies/μL . The classifier ζ was defined as 1 ˗ c . d . f . and describes the probability that the true concentration is greater than zero copies/μL . We excluded samples from further analysis if the ζ value for the RPP30 assay was below 0 . 95 . A specimen was considered positive for ddPCR when the lower 95% confidence interval of the load estimate was above zero , which is considered to optimize specificity as described by Roberts et al . 2013 [20] . Samples with over 99% droplet occupancy were considered too concentrated for accurate quantitation as they potentially contravened assumptions of the Poisson distribution . While binary diagnostic classification is possible , these specimens were retested at a 1 in 100 dilution for quantitation purposes . A standard curve of TcI epimastigote-stage T . cruzi parasites diluted 10-fold from 104–10−3 parasites/mL was repeat-tested to assess the reproducibility of the ddPCR test . Each dilution in the series was tested in five technical replicate wells on a single plate to determine within-assay variance and then in a single well on five different plates to assess between-plate variability . The analytical limit of detection was determined as the concentration of the lowest step of this series at which all 10 replicates tested positive . Due to the known variability in copy number of the TcSTR region between DTUs [33] , we also tested dilution series of TcII parasites to ensure the assay would not be affected by the differing population genetics of T . cruzi . A dilution series of the TcI trypomastigote-stage was also tested to ensure that the blood stage of the parasite was detectable . To estimate the correlation in quantitation between qPCR and ddPCR , the concentration estimates from both platforms were tested using the Pearson correlation coefficient . The number of TcSTR copies per parasite was determined using ddPCR . The number of parasites per extraction was 0 . 3 parasites per mL , as 300 μL of parasite culture was used in the extraction protocol and downstream analysis . The number of copies per μL in the ddPCR reaction was also multiplied by four to account for the dilution of DNA in the reaction and was used to quantify the number of TcSTR copies in each dilution point of the standard curve material . A total of 192 samples were included in the analysis . Blood samples were collected and stored in 6 M guanidine hydrochloride and 0 . 2 M EDTA buffer , pH 8 , as described elsewhere [9] . Multiplex qPCR for the detection of satellite DNA of T . cruzi was performed as reported elsewhere [34] . The parasitic load was measured by qPCR as parasites per mL according to Moreira et al . 2013 [29] . A total of 106 positive samples that were previously characterized by qPCR , indirect immunofluorescence assays , an indirect hemagglutination test and ELISA as positive and 86 negative samples were also characterized by the same techniques . Our samples included those from patients in the acute phase of infection ( n = 11; 10 . 4% ) . A suspected acute case was defined as an individual with >7 days of fever accompanied or not by hepatomegaly or splenomegaly . A patient was considered to have acute Chagas disease if in addition to testing positive in parasitological tests ( strout , micro-strout , blood thick smear or hemoculture ) , positive results were obtained for two serological tests over the course of the following weeks . The average patient age was 49 years ( from 5 to 61 years old ) . Our samples also included those from patients in the chronic phase of infection ( n = 95; 89 . 6% ) . These individuals did not fulfil the criteria for acute phase infection but had clinical or epidemiological suspicion of Chagas disease . The average age of these patients was 49 years ( from 26 to 88 years old ) . These patients were confirmed as positive for T . cruzi infection when they tested positive to two serological tests ( IFA , ELISA and/or HAI ) . They were then classified as chronic undetermined ( when no evidence of signs or symptoms of heart complications were evinced ) or chronic determined in all other cases . Among the set of 95 chronic patients , a total of 30 patients were symptomatic ( determined ) and the other 65 patients were asymptomatic ( undetermined ) . Descriptive analyzes were performed to report the frequencies of infection detected by qPCR and ddPCR , considering the qualitative result ( positive/negative ) for each test with respect to the total number of samples evaluated ( n = 192 ) . The agreement between the results of the tests was evaluated through the calculation of the coefficient kappa . Sensitivity , specificity and positive and negative predictive values were determined , as indicators of the operative characteristics of the ddPCR , taking as a reference test the qPCR , previously identified as being efficient for the detection of infection [14] . A 95% confidence interval ( 95% CI ) using the analytical method ( due to the dichotomous nature of the variables analyzed ) , was calculated for the main events of interest . A value of P<0 . 05 was considered significant . The statistical analyzes were performed using the software STATA 12 . 0 . Finally , to test the association between parasitic loads using both platforms ( qPCR vs . ddPCR ) , a graphic of Bland–Altman was generated using Microsoft Office Excel 2016 software . For this , the differences between the loads calculated by ddPCR and qPCR were determined , and these two values were averaged for each of the samples . Bias was then calculated by means of the average of the differences for all samples , as well as their respective standard deviations . The upper and lower limits of the deviation were calculated using the bias +/˗ 1 . 96 multiplied by the standard deviation . For the statistical tests , normality tests ( Shapiro–Wilk ) were initially carried out to determine the distribution of the data . Based on their distribution , the correlations between age and parasitic loads calculated by each of the evaluated methods were determined . The differences in the medians of the parasitic loads calculated by each of the methods , both general and according to the clinical phase and state of the disease were determined using a Mann–Whitney U test . All statistical analyzes were performed using the Stata 14 software ( StataCorp , 2015 . Stata Statistical Software: Release 14 . College Station , TX: StataCorp LP ) .
The thermal cycling conditions for the TcSTR ddPCR assay were optimized for this study . Fig 1 shows ( A ) the native assay conditions and ( B ) the optimised assay conditions across a standard curve ( wells A–G ) and one no-template control well ( well H ) . In the diagnostic target ( TcSTR ) channel ( FAM fluorophore ) , there was a small increase in the mean negative population amplitude under optimized cycling conditions compared with the native cycling conditions ( 1599 versus 1554 , p = 0 . 01 ) . There was a large increase in mean amplitude of the positive droplets under the optimized cycling conditions compared with the native ones ( 8255 versus 5027 , p = 0 . 002 ) . The ratio of positive-to-negative droplet amplitude also increased from 3 . 2 to 5 . 2 ( p = 0 . 002 ) . No-template control wells were negative for both channels under optimized and native thermocycling conditions . There was no difference in the endogenous control channel HsRPP30 ( HEX fluorophore ) negative droplet amplitude between optimised and native conditions ( 1161 versus 1146 , p = 0 . 10 ) . During repeat testing of the T . cruzi standard curve , the top standard ( 104 parasites/mL ) resulted in over 99% of the droplets being saturated . The upper limit of the dynamic range for accurate quantitation of this assay was 103 parasites/mL . Of 10 repeated readings of the 1 parasite/mL standard , all 10 were detected . Of 10 repeat readings of the 0 . 1 parasites/mL standard ( equivalent to approximately 0 . 5 TcSTR copies/μL or 2 . 5 TcSTR copies/test ) , only seven were detected . The lower dilutions were also not detected in all repeats; therefore , the limit of detection is 1 parasite/mL , measured as 5 TcSTR copies/μL or 20 TcSTR copies/test . The TcI epimastigotes tested had a mean of 9 , 286 ( range: 8 , 768–10 , 113 ) TcSTR copies per parasite . The Pearson correlation coefficient between TcSTR copies/μL and parasites/mL in the dynamic range of the dilution series was 0 . 987 . The efficiency of the assay was 98% ( Fig 2 ) . The correlation between the qPCR cycle threshold and the ddPCR load estimate was excellent in the standard curve . Upon testing of different developmental stages and different lineages , both TcII and TcI trypomastigote curves also performed equally well as the TcI epimastigote curve ( Fig 3 ) . The ddPCR between-plate coefficient of variation ( CV ) across the whole series was 101% , and the CV at 1 parasite/mL was much higher than at 1000 parasites/mL ( 200% compared with 53% ) . The within-plate CV was 13% , and the CV at 1 parasite/mL was much higher than at 1000 parasites/mL ( 30% compared with 4% ) . When comparing the within-plate CV between qPCR and ddPCR , the CV of ddPCR across the dilution series did not differ between the platforms ( qPCR CV = 18 . 6% versus ddPCR CV = 13 . 0% ) . Also , the ddPCR CV was not significantly lower than the CV of qPCR at the lower end of the dilution series ( Fig 4 ) . The frequency of positives was the same for the two tests being evaluated , at 55 . 2% [n = 106; 95% CI = 47 . 9–62 . 4] . The concordance between the results was absolute ( 100% ) and was supported by a perfect kappa coefficient ( 1 . 0000 ) . Analysis of the operational characteristics of the ddPCR test revealed great performance ( 100% in all cases ) in detecting T . cruzi DNA ( Table 2 ) , when compared with the results of the qPCR test included as a reference . For all clinical phases and totals there was absolute agreement between the tests . All positive samples were correctly classified , as were the negative samples . Therefore , all values accurately classified the samples by clinical stage ( 100% ) . The age of the patients was distributed normally ( p = 0 . 1039 ) , while the load ( copies per μL ) found by ddPCR was not distributed normally ( p = 0 . 0000 ) . Spearman correlation coefficient was determined and there was no correlation between the variables evaluated ( age vs . parasitic load; Spearman's rho = 0 . 0012 , p = 0 . 9900 ) , and no correlation was found between the variables age and copies/μL quantified by qPCR ( Spearman's rho = 0 . 0727 , p = 0 . 4586 ) . Finally , when evaluating the differences in median parasitic loads calculated by ddPCR between the two clinical phases , no statistically significant difference was found ( p = 0 . 8399 ) . Likewise , no statistically significant differences were found for the median parasitic loads calculated by qPCR between the phases ( p = 0 . 9546 ) . When evaluating the association by state with chronic infection cases , no statistically significant differences were found either by ddPCR or by qPCR ( p = 0 . 5643 and p = 0 . 0567 , respectively ) . A Bland–Altman graph was generated using Microsoft Excel 2016 software . No differences were observed in quantitation between the techniques when the parasitic loads were low . However , as the parasitic load increased , differences in the quantification between the two techniques began to become evident since ddPCR underestimates the actual parasitic load ( Fig 5 ) .
Chagas disease still presents a serious challenge to public health in Latin America and has increased in profile due to the immigration of infected individuals into non-endemic countries [35] . One of the challenges in controlling the disease is its diagnosis due to the stark contrast between the clinical signs of disease and the parasitic load between the acute and chronic phases of the disease [5] . Parasitological , serological methods and most recently molecular methods have been deployed to circumvent these challenges [10 , 15 , 36] . Due to the theoretical advantages of ddPCR outlined in the introduction , it is an attractive option for the detection and quantitation of low copy targets . To optimize the performance of the assay on the ddPCR platform , we extended the number of cycles to 50 and increased the duration of both the dissociation and annealing/extension stages of thermocycling . This was successful in that it improved separation between positive and negative droplets , and vastly improved the ease of setting appropriate thresholds in this assay . One risk associated with increased cycle number is non-specific amplification , resulting in a loss in specificity . In this case the increase from 40 to 50 cycles had no effect on analytical specificity ( as evidenced by consistent zero readings during repeat testing of the no-template control samples ) and the excellent clinical specificity of the assay . Cycle number has also been increased to optimize assays in other contexts , such as the detection of genetically modified organisms [37] , the detection of mutations in AH1N1-infected individuals [38] and the detection of foodborne pathogens in soft cheese [39] . One limitation of increasing the cycle number is the concomitant increase in the sample testing time; this may be problematic in high-throughput settings [31 , 40] . We found the range of accurate quantitation of this assay to be between 100−103 parasites/mL , equivalent to approximately 5–5000 copies/μL . Typically , this would be considered to be within acceptable limits of performance for research purposes . The limit of detection of ddPCR was acceptable , and assay quantitation correlated well with parasitological findings ( Pearson correlation coefficient: 0 . 987; efficiency: 98%; Fig 2 ) . However , compared with qPCR , we could not demonstrate the theoretical improvement in precision at low loads that we expected . The CV was not significantly different between ddPCR and qPCR at any concentration tested [14] . However , it was interesting to observe that the within-plate CV of qPCR was higher at 1 parasite/mL than at 10 parasites/mL . This might be explained because at low concentration molecular tests tend to be unstable . This has been already reported for qPCR [14] . When comparing the limit of detection with that of qPCR , the limit of detection reported for this assay in the qPCR format was 0 . 46 parasites/mL [6] , whereas in the ddPCR format , only 7/10 repeats were successfully detected at that level . This has been reported for other pathogens , where qPCR has been shown to be more sensitive , e . g . in Cytomegalovirus and Xanthomonas citri subsp . citri [22 , 41] . There are a number of potential reasons for this , many of which are outlined by Hindson et al . [19] . The assay dynamics could be more suited to the qPCR well format . Also , the TcSTR is a variable region , as recently demonstrated , and its lack of stability could affect the sensitivity of a high-resolution test such as ddPCR [42] . One interesting finding was the difference between the dynamic range when testing epimastigotes versus trypomastigotes . To date , no studies have conducted standard curves with the human-infective stage of T . cruzi . Our results showed that the standard curve with trypomastigotes showed one order of magnitude more DNA than epimastigotes ( Fig 3 ) . This could be explained by the cell size of trypomastigotes as they may have double or triple the cell surface of epimastigotes and subsequently have more DNA content [43] . Also , it might be related to the genomic architecture of T . cruzi as it was recently shown that: i ) the DNA transcripts change drastically between cell stages [44 , 45] , and ii ) trypomastigotes present more telomeres and DNA content than epimastigotes using next generation sequencing [46 , 47] . However , these assumptions need to be validated by conducting more experiments including the testing of metacyclic trypomastigotes from other DTUs . We are aware that one limitation of our study is that a strict analytical validation was not conducted as stated in international guidelines for clinical laboratories . However , this has not been done even for the qPCR or PCR platforms . We used the same pipeline of validation established for ddPCR platforms already published and it must be kept in mind that PCR platforms are not currently considered clinical diagnostic tools for Chagas disease . When evaluating the diagnostic performance of this assay , the concordance between qPCR and ddPCR was found to be excellent ( kappa = 1 ) . With perfect agreement between the positives and negatives for each of the tests , the operational characteristics showed 100% in all cases ( Table 2 ) . The Bland–Altman plot also showed a good degree of association between the parasitic loads for both platforms ( Fig 5 ) . Although at high concentration , the ddPCR platform became saturated and qPCR showed a better performance , explaining the outliers from the graph . In general , this shows that in this sample set , despite not offering a technical performance benefit over qPCR , the use of this assay on the ddPCR platform may be as suitable as qPCR for the detection of T . cruzi . However , one study reported median parasitemia values in chronic Chagas disease patients from Argentina and Colombia to be 1 . 93 and 2 . 31 parasite equivalents/mL , respectively . Based on our present evaluation , the ddPCR assay should be able to reliably detect samples with these median loads but may fail on occasion with lower median parasitemia levels . More concerningly , the median load of parasitemia for Brazilian patients was 0 . 1 parasite equivalents/mL , which is below the reproducible limit of detection for the ddPCR assay [29] . A recent study of patients in the chronic phase showed a sensitivity of 64 . 2% and specificity of 97 . 1% for qPCR [6 , 48] , suggesting it too will not reliably detect all infections in chronic patients , resulting in a loss of sensitivity . Finally , the cost of ddPCR is known to be approximately double the price of qPCR , therefore without a significant performance improvement , it is unlikely that this constitutes a realistic public health surveillance tool . However , the primary utility of the ddPCR platform on which to run this assay is likely to be as a research tool . In our study , the ddPCR assay was not efficient at high concentrations of T . cruzi ( Fig 1 ) . This impedes its utility in acute Chagas disease patients , where parasitemia levels tend to be high and would suggest that DNA samples would need to be diluted prior to application on a ddPCR platform . A possible solution might be exploring other molecular targets across the T . cruzi genome that may have a lower number of copies per cell; however , this could affect the sensitivity . Previously , some authors suggested the use of 18S , 24S , cytochrome oxidase II ( COII ) or the spliced leader of miniexon gene for the molecular diagnosis of T . cruzi . However , in an international consensus , these markers did not show satisfactory results in terms of sensitivity [13] . New experiments would need to be performed with these molecular targets under a ddPCR platform to determine their potential use . Additionally , in a recent analytical validation report for T . cruzi , the kinetoplast ( kDNA ) showed interesting results in terms of sensitivity that were improved for TcSTR [14] . We initially tested the kDNA reported by Ramírez et al . 2015 [14] and found oversaturation of the droplets preventing adequate discrimination of positive and negative droplets as with TcSTR . This might reflect the molecular nature of kDNA ( discoid DNA with minicircles and maxicircles ) . Therefore , we did not pursue this marker . In Colombia , Trypanosoma rangeli is a frequent parasite and kDNA presents strong cross-reactivity with this protozoan [49] . Therefore , we concluded that TcSTR was the most suitable approach to initiate the analytical validation of a ddPCR platform . However , maybe an optimized primer or probe design could improve the analytical characteristics of ddPCR and additional experiments are needed to identify appropriate molecular candidates . In several studies , ddPCR performed much better than qPCR for the detection of pathogen DNA ( for example , C . trachomatis , Salmonella , E . coli , C . jejuni , cytomegalovirus and HIV ) [20–24] . When applied to the detection of Plasmodium species , this tool was able to quantify low parasitemia in subpatent infections , as well as determine Plasmodium species in cases of low parasitemia [50] . In the case of Cryptosporidium , the precision of ddPCR was consistently higher compared with qPCR but decreased as the DNA concentration decreased [27] . We did not identify the same improvement in the detection of T . cruzi in blood samples after importing the assay to the ddPCR platform . However , this corresponds a pioneer study towards the establishment of an efficient protocol to employ ddPCR as a test to detect T . cruzi DNA in blood samples . As mentioned before , novel targets and better primer design might potentially improve the performance of the assay . There are a number of limitations to our study . The samples used for the evaluation of clinical performance did not incorporate samples from other clinical settings outside of Colombia . This is important as the median parasitic load is higher in Colombian patients , and therefore the sensitivity may be reduced in patients from other regions . Also , the clinical stages , age and patient characteristics could impact on the results of ddPCR . In our case , we observed that T . cruzi DNA could be detected in acute and chronic patients by ddPCR with no statistical significance on the parasitic load in terms of age or clinical stage . However , verification is required in a novel group of patients including , for example , children and congenital cases . Additionally , the variation in TcSTR copy number between DTUs and even within DTUs may cause the results to be variable in clinical settings; a fixed copy number target would be preferable for the purposes of quantitation . In conclusion , we report the first evaluation of the suitability of a ddPCR platform to detect T . cruzi DNA in blood samples . The analytical and diagnostic performance of the ddPCR assay evaluated here does not support its routine usage over qPCR . However , it shows several advantages such as quantitation of DNA without the need for calibration curves thereby allowing for exact and reference values to be obtained per sample ( higher-order reference measurement method ) as has been suggested for various viral diseases . A range of samples from a larger geographical area should be tested and a future international consensus should be considered as was conducted for qPCR to comprehensively determine the potential use of ddPCR to detect and quantify T . cruzi in clinical samples . | Chagas disease is a complex pathology caused by the parasite Trypanosoma cruzi . This disease affects more than 8 million people in Latin America and is considered a serious public health problem . The diagnosis of the disease is challenging due to the natural course of infection and therefore molecular tools such as qPCR have been deployed to detect parasite DNA in infected individuals . However , qPCR is not suitable for diagnosis in the chronic phase of the disease and new diagnostic tools are needed . Digital droplet ( dd ) PCR is a next-generation PCR technology that enables the absolute quantitation of nucleic acids . This technology has been successfully applied as a diagnostic tool in bacteria and viruses , but few reports exist for parasites . Herein , we imported a previously validated qPCR assay to the ddPCR platform . Our results show complete agreement in terms of diagnostic operative features but a higher limit of detection . Overall , ddPCR did not offer significant diagnostic benefits compared with qPCR . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"tropical",
"diseases",
"microbiology",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoan",
"life",
"cycles",
"organisms",
"developmental",
"biology",
"trypomastigotes",
"protozoans",
"neglected",
"tropical",
"diseases",... | 2018 | Evaluation of the analytical and diagnostic performance of a digital droplet polymerase chain reaction (ddPCR) assay to detect Trypanosoma cruzi DNA in blood samples |
Anatomically based procedures to ablate atrial fibrillation ( AF ) are often successful in terminating paroxysmal AF . However , the ability to terminate persistent AF remains disappointing . New mechanistic approaches use multiple-electrode basket catheter mapping to localize and target AF drivers in the form of rotors but significant concerns remain about their accuracy . We aimed to evaluate how electrode-endocardium distance , far-field sources and inter-electrode distance affect the accuracy of localizing rotors . Sustained rotor activation of the atria was simulated numerically and mapped using a virtual basket catheter with varying electrode densities placed at different positions within the atrial cavity . Unipolar electrograms were calculated on the entire endocardial surface and at each of the electrodes . Rotors were tracked on the interpolated basket phase maps and compared with the respective atrial voltage and endocardial phase maps , which served as references . Rotor detection by the basket maps varied between 35–94% of the simulation time , depending on the basket’s position and the electrode-to-endocardial wall distance . However , two different types of phantom rotors appeared also on the basket maps . The first type was due to the far-field sources and the second type was due to interpolation between the electrodes; increasing electrode density decreased the incidence of the second but not the first type of phantom rotors . In the simulations study , basket catheter-based phase mapping detected rotors even when the basket was not in full contact with the endocardial wall , but always generated a number of phantom rotors in the presence of only a single real rotor , which would be the desired ablation target . Phantom rotors may mislead and contribute to failure in AF ablation procedures .
Atrial fibrillation ( AF ) is the most common cardiac arrhythmia seen in clinical practice and is the most important cause of embolic stroke[1 , 2] . Recently catheter ablation has been recommended as a first-line treatment for AF termination[3] . Traditionally ablation procedures aimed at terminating AF have been primarily focused on isolating the pulmonary veins ( PVs ) [4–6] often complemented by linear ablation of the posterior left atrium ( LA ) [7] . Recent approaches based on mapping the electrical activity during AF take into account the underlying mechanism and target the AF drivers[8 , 9] , as has been demonstrated by high-resolution optical mapping of AF in animal models[10–13] and explanted human hearts[14] . Unfortunately , clinical mapping approaches are limited to the use of low resolution multi-electrode systems ( contact and non-contact ) [8 , 15–21] , but the use of panoramic contact multi-electrode basket catheters to map the atria in search for AF drivers enabled >80% success rates in some studies compared to 20–50% obtained by conventional ablation[1 , 15 , 22] . However , whether rotors are AF drivers remains controversial[18 , 19 , 23 , 24] and the use of multi-electrode mapping approaches to target rotors needs further studies to validate their accuracy and applicability in the clinic . We surmised that several factors may limit the accuracy of basket catheter-based phase mapping in localizing rotors , including: ( 1 ) the electrode-endocardium distance; ( 2 ) the effects of distant electrical sources and ( 3 ) inter-electrode interpolation . Here we use computer simulations to analyze how those factors affect the accuracy of basket catheters in mapping AF-like electrical activity and detecting rotors .
We used a recently developed realistic 3D model of the human atria[25] that included heterogeneity at cell , tissue and organ scale . It comprises the following anatomical regions with their respective fiber orientation: right atrium ( RA ) , left atrium ( LA ) , crista terminalis ( CT ) , fossa ovalis ( FO ) , pulmonary veins ( PV ) , Bachman’s bundle ( BB ) , divided into left ( BBLA ) and right ( BBRA ) sides , pectinate muscles ( PM ) , isthmus ( IST ) , sinoatrial node ( SAN ) , coronary sinus ( CS ) , mitral valve ring ( MVR ) , tricuspid valve ring ( TVR ) , right atrial appendage ( RAA ) , left atrial appendage ( LAA ) , superior vena cava ( SVC ) and inferior vena cava ( IVC ) . This computational finite element model is composed of linear hexahedral elements with a regular spatial resolution of 300 μm , and wall thickness between 600 and 900 μm ( 754893 nodes and 515010 elements ) . See [25] for additional details . The electrical activity in our model was solved by the monodomain formalism , and the cellular ion kinetics by the Courtemanche-Ramírez-Nattel ionic model[26] . In order to reproduce transmembrane potential ( Vm ) of experimentally observed heterogeneity in action potentials ( APs ) morphology and duration in different regions of the atria[27–31] , the maximum conductance of three ionic currents ( Ito , ICaL and IKr ) was adjusted as described previously[25 , 32–35] ( see Table 1 ) . This procedure yielded nine cellular models ( RA/PM , LA , CT/BBRA , BBLA , PV , MVR , TVR , RAA and LAA ) whose APs are depicted in Fig 1A ( top ) ; AP durations ( APD ) to 90% and 95% repolarization ( APD90 and APD95 ) are shown in Table 1 . APs were recorded after 1 minute of stimulation at a basic cycle length ( BCL ) of 1000 ms; pulse amplitude and duration were 28 pA/pF and 2 ms , respectively . The APD variation among regions was similar to experimental observations[27–29]: APD was longer in the CT region than in the RA region , where it was longer than in the TVR , RAA and LA regions; and APD was shorter in the PV region than in the LA region . Then the nine cellular models corresponding to healthy conditions were assigned to the nodes in the 3D geometrical model , following the distribution shown in Fig 1B . Tissue conductivities in each region were tuned as in[25] to match the activation sequences to experimental data[36] ( see S1 Fig ) . Atrial electrical remodeling corresponding to chronic AF ( cAF ) was also introduced through the variation of the maximum conductances of Ito , ICaL , IK1 , IKur and IKs ( see Table 2 ) [37–42] , similarly to other computational studies[33 , 43–47] . Remodeling in the RA ( RA , CT/BBRA , TVR and RAA regional models ) was different to that in the LA ( LA , BBLA , MVR , LAA and PV regional models ) , according to experimental data reported in the literature ( resulting APs depicted in Fig 1A bottom ) . APD shortening was within the range of the available experimental observations[38 , 40] ( Table 3 ) . The maximum dispersion of APD90 due to heterogeneity in our model was 57 ms in cAF , compared to 125 ms under control conditions . The nine remodeled cellular models were considered across the atria as for control conditions ( Fig 1B ) , with 15% reduced intracellular conductivity to account for gap junctional remodeling[48 , 49] as in other simulation studies[44 , 47] . Then we applied 21 stimuli ( BCL 500 ms , amplitude 28 pA/pF , duration 2 ms ) to the SAN region to stabilize models in neighboring regions . The electrical and gap junctional remodeling produced a reduction of 17% in the conduction velocity with respect to control , consistent with experimental observations[49] . To generate reentrant activity , after the 21 stabilization pulses , we paced the CS using a continuous high frequency train of ectopic foci ( cycle length = 110 ms ) and the simulation was run for an additional 12 seconds , in which the simulated activity was driven by 2 stable sources: the pacing train near the CS and a stable rotor near the CT . Stimulation from the SAN was stopped upon starting the ectopic pacing from the CS because , in test simulations , when SAN stimulation was maintained , its discharge propagation was overridden by the faster ectopy rate ( 110 ms vs 500 ms ) and differences in the propagation patterns were negligible . Unipolar EGMs inside the atrial cavity and on its endocardial boundaries were computed as extracellular potentials , with a temporal resolution of 1 ms , in a whole atrial-torso model developed previously[25] . The torso model developed in[25] was re-meshed in order to improve the spatial resolution at: 1 ) the endocardium-blood interface to enhance the accuracy when detecting the rotor’s tip trajectory; 2 ) the atrial blood to introduce the EGMs at the basket electrodes in several positions . Accordingly , the resulting torso mesh had 254976 nodes and 1554255 tetrahedral elements with a spatial resolution ranging approximately from 0 . 5 mm on the atrial endocardium-blood interface to 5 . 8 mm on the torso surface . The number of nodes belonging to the endocardium-blood interface were 25175 in the RA and 24819 in the LA . The extracellular potentials were computed by the bidomain formalism in two steps[50] . By assuming equal anisotropy ratios for the intracellular , Di , and extracellular , De , conductance tensors ( De = λDi ) , the bidomain equations can be decoupled into an equation describing the changes in the transmembrane potential , Vm , and another equation describing the extracellular potential , Ve , in the heart domain[51]: ∇∙ ( D∇Vm ) =Cm∙∂Vm∂t+IioninΩH ( 1 ) ∇∙ ( D∇Ve ) =−11+λ∇∙ ( D∇Vm ) inΩH , ( 2 ) where D=λ1+λDi is the equivalent conductivity tensor , Vm is the transmembrane potential , Ve is the extracellular potential , Iion is the transmembrane ionic current that depends on the cellular model , Cm is the membrane capacitance and ΩH is the heart domain . Eqs ( 1 ) and ( 2 ) are subjected to the following boundary conditions in ΩH: n∙ ( D∇Vm ) =0on∂ΩH ( 3 ) n∙ ( D∇Ve ) =0on∂ΩH , ( 4 ) where n is the outward normal to ∂ΩH . Eqs ( 1 ) and ( 3 ) allow for solving the Vm in the cardiac tissue , whereas Eq ( 2 ) and ( 4 ) recast the Ve in the heart tissue after Vm has been calculated . Note that boundary conditions ( 3 ) and ( 4 ) consider the heart to be immersed in a non-conducting bath . To calculate the EGMs inside the heart cavity and on the basket electrode points , we need to place the heart within the torso and solve for the extracellular potential in the entire domain ( the heart ΩH and the torso ΩT outside of the heart ) . Therefore , we define our 3D space problem to include the governing equations for the solid conductor associated with the torso , and modify accordingly the boundary conditions at the heart/torso interface , i . e . , ∂ΩH . Under the hypothesis of equal anisotropy ratio for Di and De , the extracellular potential in the domain ΩH∪ΩT , after obtaining Vm as solution of Eq ( 1 ) and ( 3 ) , is found as the solution of the following Laplace Equation: ∇∙ ( DT∇VT ) =0inΩT ( 5 ) where VT and DT are the extracellular potential and heterogeneous conductance tensor in the torso , respectively . Eq ( 5 ) is subjected to the following boundary and continuity conditions: Ve=VTin∂ΩH ( 6 ) n∙ ( D∇VT ) =0in∂ΩT ( 7 ) where ∂ΩT is the torso-air non-flux boundary . Then , VT is the EGM at either the basket electrodes positions or at the endocardial surface . Next , endocardial and cavity EGMs ( i . e , the VT ) were bandpass filtered ( 7–10 Hz ) in order to allow rotor tip tracking similarly to[52] . Phase maps were calculated from the EGMs during the last 11 s of the simulation , since the reentrant activity during the first second was not stable . In the custom-made software routines implemented in MATLAB ( MathWorks , Natick , MA ) , we applied the Hilbert transform ( HT ) to the filtered EGMs ( EGMf ( t ) , Eq 8 ) , as in prior studies[11 , 53 , 54] and computed the instantaneous phase θ , whose values ranged from -π to π radians ( Eq 9 ) . We excluded the first and last 500 ms of the signals to avoid filtering and transformation artefacts , resulting in phase maps as frames of movies of 10-s long . Finally the phase singularity points ( PSs ) , where all phases converge , were computed to track the rotor’s trajectory on the endocardium-blood interface [55–58] when: ( 10 ) where r is the closed curve surrounding the singularity point at the center of the reentry . To approximate Eq ( 10 ) and automatically localize PSs , we adapted the method proposed by Rogers[59] . Accordingly , the endocardium-blood interface corresponds to a 3D surface mesh of triangular elements and each node in the mesh has a phase value . For each element in the mesh and for all simulation time steps we computed the spatial variations of phase among its nodes ( space gradients ) . For those elements around which all phases converged , the summed phase variation along a closed surrounding path was ±2π . Those elements were designated as PSs and painted in white superimposed on the phase maps ( see S1 Video panel B ) . The automatic PS detection algorithm was implemented in MATLAB . A similar procedure was performed on the basket phase maps . The closed path integral was computed along the edges of each triangular element ( with a perimeter of approximately 0 . 5 mm in case of the 2D projections of the basket , and 2 mm in case of the endocardial-blood surface ) . Rotors were defined as an excitatory wave pivoting around a PS for at least one cycle of rotation and were visually detected in phase map movies ( see S1–S4 Videos ) . A virtual intracardiac 64-pole mapping basket catheter formed by 8 splines ( A-H ) , each containing 8 electrodes ( 1–8 ) , was placed in three different positions in the RA of our atrial model: close to the SVC , CT and CS ( Fig 2 ) . The diameter of the 8×8-electrode basket modelled was 31 mm , corresponding to the smallest Constellation mapping catheter ( Boston Scientific ) . Electrodes were equidistant along the splines with an inter-electrode distance of 4 . 8 mm . The distances between electrodes at neighboring splines varied between 5 . 4 mm for the electrodes near the poles and 11 . 7 mm for the electrodes near the equator ( see S2A Fig ) . EGMs were computed at the 64 electrodes’ coordinates and linearly interpolated on 57600 points on a periodic 2D projection of the basket ( Fig 2A–2C right column ) . The same signal processing explained above was applied to the basket’s EGMs , in order to obtain the phase maps and PSs detection . The distribution of the distance ( d ) between each electrode in the basket and the closest point on the endocardial surface differed for the three basket positions within the atria , but the number of electrodes located at d≤0 . 5 cm and 0 . 5<d≤1 cm was approximately similar ( Fig 2A–2C right column ) . In the first case the percentage of electrodes was 53% , 58% and 50% , and in the second case 41% , 34% and 42% , for the SCV , CT and CS positions , respectively . Finally , to analyze the effect of electrode density , we varied the number of electrodes in the basket between 4×6 and 16×16 for each of the three positions ( see S2 Fig ) . The mono-domain formulation ( Eqs 1–4 ) was solved using the operator splitting numerical scheme with ELVIRA software[60] with a constant time step dt = 0 . 01 ms . Simulation of 12 s of atrial activity took 37 hours on a computing node with eight 6-core AMD Opteron Processors 6234 clocked at 2 . 4 GHz . The approximation of the bidomain formulation ( Eqs 5–7 ) , the phase maps ( Eqs 8–9 ) and the PSs detection ( Eq 10 ) were computed with custom-made software routines in MATLAB ( MathWorks , Natick , MA ) .
Application of a sustained high frequency train of stimuli to the RA close to the CS led to a complex propagation pattern maintained by a stable rotor on the crista terminalis ( CT rotor ) accompanied by a distal rotor wave extension ( RWE ) reentry around the inferior vena cava ( IVC; Panel A in S1 Video ) . The CT rotor migrated back and forth between the superior vena cava ( SVC ) and the IVC along the CT , while its RWE persistently collided and merged with the wavefront generated by the CS stimuli . Propagation in the LA was more regular . The LA was repeatedly excited by two wave fronts from the RA: one moving toward the roof through Bachmann´s bundle ( BB ) and spreading to the posterior wall and the other moving toward the inferior wall through the foramen ovale ( FO ) limb and extending to the posterior wall . Both wave fronts collided between the posterior and inferior walls , with the precise location of the collision line varying throughout the simulation . Fig 3A is a snapshot at 5125 ms showing the CT rotor ( white arrow ) , the RWE ( dashed white arrow ) , the direction of the propagation in the LA ( dotted white arrows ) , the wave front corresponding to the high frequency train of stimuli applied near the CS ( dashed orange arrows ) , the collision between the CS stimuli wave front and the RWE ( black line close to the FO ) and the collision in the LA ( black line between the posterior and inferior walls ) . The EGMs and their corresponding phase maps were calculated at the endocardium-blood interface . The trajectory described by the CT rotor tip , detected through the PSs calculations , is depicted in white superimposed on the phase maps ( Panel B in S1 Video ) . All the events described above for the propagation maps were also identified on the phase maps , as shown in Fig 3B . Furthermore , not only the CT rotor but also the RWE circulating around the IVC appeared as a rotor whose tip was in the middle of the orifice . The rotational activation of the tissue can be observed in the EGMs shown in Fig 3C . The EGMs on the left correspond to red points 1–5 ( CT rotor ) in Fig 3A , while the EGMs on the right correspond to red points 6–10 ( RWE ) . Their corresponding phases are shown in S3A Fig . The amplitude of the EGMs depends on the dipole strength and the source-to-electrode distance ( see Fig 4 ) . For all three different basket positions ( SVC , CT and CS ) we analyzed the effect of the electrode-endocardium distance ( d ) at each electrode location by computing EGMs for all 64 electrode coordinates to build phase maps . We also evaluated the RA and the meandering area coverages for each basket position . First , the RA coverage was defined as the percentage of endocardium at a distance d ≤ 0 . 5 cm ( red ) or 0 . 5 < d ≤ 1 . 0 cm ( green ) from at least one electrode of the basket ( Fig 4A ) over the whole atrial tissue . RA coverage was similar for the three basket positions ( ~20% for d ≤ 0 . 5 cm and ~45% for 0 . 5 < d ≤ 1 . 0 cm , as shown in Fig 4B ) . Second , the meandering area coverage was defined as the percentage of CT rotor meandering area ( black area in panel A , corresponding to the region of the endocardium comprising the rotor trajectory ) superimposed on the RA coverage ( red or green area in panel A ) over the whole CT rotor meandering area . Meandering area coverage was ~60% for d ≤ 0 . 5 cm and ~40% for 0 . 5 < d ≤ 1 . 0 cm for the SVC and CT positions ( Fig 4A , top and middle , and Fig 4C ) , while it remained completely uncovered for the CS position ( Fig 4A , bottom , and Fig 4C ) . As an example of the variation of amplitude regarding the distance to the tissue , the traces on the right-hand side in panel A of Fig 4 shows EGMs on electrodes set at d ≤ 0 . 5 cm ( Red traces; A3 , A5 and F8 for the SVC , CT and CS positions , respectively ) and at 0 . 5 < d ≤ 1 . 0 cm ( Green traces; B7 , C1 and A1 for the SVC , CT and CS positions , respectively ) . The maximum root mean square value ( Vrms ) was 0 . 74 mV among all the EGMs corresponding to the electrodes set at d ≤ 0 . 5 cm , while it was 0 . 33 mV for the electrodes set at 0 . 5 < d ≤ 1 . 0 cm . Interestingly , a relatively large variation on the nearest EGMs amplitude at electrodes A3 and A5 is observed in the SVC and CT position which could be related to rotor meandering . The effects of the 3 different positioning of the basket within the RA on the characteristics of its rotor mapping are illustrated in Figs 5–7 . Surprisingly , in addition to the CT with its RWE on the atrial surface ( see Fig 3 ) , the basket phase maps also show phantom PSs at various locations that have no corresponding reentrant AP on the atria . We classified the false PSs either as an imaginary phase singularity ( IMPS ) , when activation occurred sequentially in the surrounding electrodes , or as a false interpolation phase singularity ( FIPS ) , when the electrodes surrounding the singularity did not register the activation sequentially . FIPSs appear because of the inter-electrode interpolation of the EGMs prior to the computation of the phase maps ( see below ) . For the three basket positions tested , a comparison between the PSs locations on the phase maps and the electrode-to-endocardial surface distance maps ( Figs 2 and 4 ) shows that IMPSs and FIPSs appear at various distances , including d≤0 . 5 cm , suggesting that near- and far-field sources have an influence on the generation of PSs . Starting from the SVC position in Fig 5 , panel A illustrates the location of the basket at the high RA ( top row ) , the phase maps on the endocardium-blood interface ( middle row ) and the phase map corresponding to the basket recordings ( bottom ) at t = 10000 ms . White colored PS trajectories are superimposed on the phase maps . The basket phase map was visually more like the endocardium-blood interface map at splines A-B-C-D than at splines E-F-G-H , where a higher number of electrodes were at d>0 . 5 cm from the tissue ( see Figs 2 and 4 ) . The CT rotor was detected 94% of the simulation time at splines A-B-C ( white clockwise arrow in Fig 5A ) . The CT rotor was not detected when it migrated closer to the IVC ( 6% of the simulation time ) . A counter-rotating wave to the CT rotor was found to rotate around the IVC and is termed in this study as a “rotor wave extension” ( RWE ) . The basket map detected the RWE mostly uninterruptedly around the IVC , near the basket’s south pole ( dashed white counterclockwise arrow ) . However , the corresponding PS appeared on the phase map only over 7% of the simulation time . Other times , the RWE was represented at the basket phase map as a wave encircling the south pole without a PS . Additionally , false PSs were also found , as follows: at splines E-F-G the basket detected a pair of imaginary PSs ( IMPSs , black arrows in Fig 5A , bottom ) during the entire simulation time ( splines near the TV orifice ) , and the basket phase map pattern was differed from the phase map pattern at the endocardium-blood interface . On the other hand , at electrodes B7-B8-C8-C7 , the basket phase map displayed false interpolation PSs ( FIPSs ) 57% of the simulation time ( grey arrow in Fig 5A bottom ) . Fig 5B depicts a selection of points ( P1 , P2 and P3 ) at the CT rotor meandering area ( top ) as well as snapshots of the basket phase maps when the CT rotor tip is located at or near each of these points . As demonstrated by the three sequential basket phase maps at 1000 , 1580 and 8600 ms , the CT rotor ( green dots ) was not detected when it moved toward the IVC , whereas the IMPSs ( red dots ) appeared for the entire simulation time ( see also S2 Video ) . Fig 5C shows the EGMs at some of the electrodes detecting the CT rotor ( C3 , B2 and B3 ) , a IMPSs ( F4 , F5 and G4 ) , the RWE ( C7 , B8 and B7 ) and FIPSs ( C8 , C7 and B8 ) . Green , red and blue arrows superimposed on the plots of the EGMs illustrate the sequential activation of the electrodes for the CT rotor , IMPSs and RWE respectively ( blue circles on the EGM plot depict the activation times ) . However , the grey superimposed arrow shows the non-sequential activation of the electrodes in case of FIPSs . At t = 10000 ms , the basket phase map ( Fig 5A , bottom ) shows the RWE and a FIPS between electrodes B7 , B8 , C7 and C8 . As demonstrated by the EGMs in Fig 5C , the counterclockwise activation corresponds to the RWE , being the FIPS an artifact due to the interpolation of the EGMs . The phases corresponding to EGMs in Fig 5C are shown in S3B Fig . The basket phase map also differed substantially from the endocardium-blood interface phase map when the basket was set at the CT position ( Fig 6 ) . At splines A-B-C , the CT rotor was detected 90% of the simulation time . During the other 10% the CT rotor had migrated closer to the SVC from which all basket electrodes were at d > 0 . 5 cm ( see Fig 6A ) . As for the RWE , its propagation was detected uninterruptedly at electrodes located at the basket’s south pole , although no corresponding PSs appeared on the phase map because the south pole ( as well as the north pole ) area was not interpolated ( see S4 Fig ) . Greater discrepancies between the endocardium and basket maps are visible at splines E-F-G , since in the CT position of the basket a higher number of electrodes corresponding to such splines were located at d > 0 . 5 cm ( see Fig 2 ) from the atrial surface . Moreover , an IMPS was noticed at splines E-F-G-H 100% of the simulation time . Finally , FIPSs appeared at electrodes B1-B2-C1-C2 and D4-D5-E4-E5 13% of the simulation time . Fig 6B shows that the CT rotor ( green points ) was not detected when it was near the SVC . In contrast the IMPS ( red points ) was always present ( see S3 Video ) . Fig 6C shows the EGMs at some electrodes detecting the CT rotor ( B6 , C6 and C5 ) , the IMPSs ( F2 , G2 and G1 ) and FIPSs ( B1 , B2 and C1 ) . Results are in accordance with those obtained for the basket in the SVC position: the activation arrived sequentially to the electrodes in case of the CT rotor and the IMPSs , but not in case of FIPSs . Phases corresponding these EGMs are shown in S3C Fig . Fig 7A illustrates the mapping when the basket was set at the CS position . The CT rotor was detected in this basket position only 35% of the simulation time at splines H-A-B between the first and second ring of electrodes in the basket ( see Fig 7A bottom ) . Detection of the CT rotor was poor because the rotor meandering area was relatively far ( d>1 cm ) from splines A-B-C of the basket in this position ( compare Fig 7A , bottom , and Fig 2 ) . The RWE was detected at electrodes B7-C7-C8-B8 61% of the simulation time . In addition , IMPSs were located along spline B ( electrodes covering the SVC orifice ) during 31% of the simulation time . And also , some FIPSs were detected 40% of the simulation time at electrodes E2-E3-E4-F2-F3-F4 and A7-A8-C7-C8 . As in the previous cases , when comparing the basket phase map to the map on the endocardium blood interface , there were some differences , especially at splines A-B-C . As shown in Fig 7B and S4 Video , neither the CT rotor ( green points ) nor the IMPSs ( red points ) were always detected . As in the other 2 basket’s positions , Fig 7C shows the EGMs at some of the electrodes detecting the CT rotor ( H1 , H2 and A1 ) , IMPSs ( B3 , C3 and B4 ) , the RWE ( B7 , B8 and A8 ) and FIPSs ( B7 , B8 and C8 ) . As in those basket’s positions , activation arrived sequentially to the electrodes in case of the CT rotor , IMPSs and RWE , and there was not sequential activation of the electrodes in case of the FIPSs . The phases corresponding to the EGMs in Fig 7C are shown in S3D Fig . The detailed analysis of basket detection for the three considered positions yielded a high detection rate ( ~90% of the simulation time ) of the real rotor , i . e . the CT rotor , when the basket was closely covering the rotor meandering area ( SVC and CT positions , as shown in Fig 4A ) . In the case of the basket at the CS position , detection of the CT rotor was poor due to the electrodes distance from the meandering area ( see Fig 4A ) . However , basket-based phase maps in our simulated setting always generated a number of phantom rotors ( IMPSs and FIPSs ) in the presence of only one real rotor ( CT rotor ) for each of the basket positions . ( It should be noted that the PS associated with the RWE around the IVC is not to be considered false due to the fact that multi-electrode recording systems will produce a PS for both anatomical and functional reentries . ) Furthermore , after computing the instantaneous location of the CT rotor on the endocardial surface along the simulation time , as well as the location of the CT rotor detected by the basket in the three different positions studied , we calculated the distance between the real and detected trajectory of the rotor at any moment of the simulations and obtained that: i ) for the SVC position this distance was between 0 . 37 and 1 . 38 cm ( median 0 . 69 cm ) ; ii ) for the CT position it was between 0 . 43 and 2 . 48 cm ( median 0 . 94 cm ) ; and iii ) for the CS position it was between 2 . 53 and 3 . 97 cm ( median 2 . 88 cm ) . This supports the fact that rotor localization is more accurate when the basket is properly located inside the atrium in close proximity to the region of the rotor . To verify the importance of far-field sources in the genesis of IMPSs , we computed the endocardial and basket maps when considering activity generated by decreasing areas of endocardial wall activity . As an example , Fig 8 shows results at t = 4935 ms for the CT position of the basket . The left most column displays the 3 spatial extensions of AP sources considered in the analysis: A1 includes the entire atrial tissue , B1 includes only tissue encompassing the CT rotor and its RWE , and C1 considers only tissue closely encompassing the CT rotor ( see S6 Fig ) ; A2 , B2 and C2 display endocardial-blood interface EGM; A3 , B3 and C3 display endocardial-blood phase maps; and A4 , B4 and C4 display the basket phase maps . When considering the whole atrial tissue ( A1 ) , the CT rotor ( white curved arrow ) and the RWE ( dashed white curved arrow ) were detected in both the endocardium-blood interface ( A3 ) and the basket phase maps ( A4 ) . Additionally , an IMPS and an imaginary extension of that rotor ( IM-RWE , hatched yellow arrows ) appeared on the basket phase map in an area close to where the endocardial-blood EGMs showed propagation around the tricuspid valve ( TV ) annulus ( blue arrows in A2 ) . As the IMPS was not detected on the endocardium-blood interface phase map ( A3 ) but was detected on the basket mapping ( A4 ) we further investigated whether IMPS are formed by far-field signals by excluding the nearby sources . When considering the sources as the activity at the CT rotor and RWE alone ( B1 ) , the EGMs in the TV area show very small voltage amplitude ( hatched blue arrows in B2 ) that gives rise to IMPSs on the endocardium-blood surface ( orange arrows in B3 ) , as well as on the basket phase maps ( hatched orange arrows in B4 ) . These orange arrows ( B3 ) indicate PSs outside of the sources region; that is , they are IMPS resulting from far field sources ( see also S6 Fig ) . A magnification of the amplitude scale of the EGMs reveals that the IMPSs in this case arise from the high sensitivity of the phase analysis to low amplitude waves far from the sources ( see S5 Video and S7 Fig ) . Finally , phase maps of sources confined to only close vicinity of the CT rotor core ( C1 ) removed the IMPSs and detected the CT rotor and its extension reentry toward the boundary of the source region ( C3-C4 ) , also termed RWE . The PS corresponding to the CT rotor remained close to its original location , shown in panels A3 and A4 . However , the RWE suffered a dramatic shift in location relative to its origin , shown in panels A3-A4 , and resided outside of the active sources region . The data in Fig 8 demonstrate that basket IMPSs were a consequence of the distal atrial tissue activation at either the CT rotor and the IVC RWE , or at the TV region . In the example provided , far-field sources interfered with the recordings when electrodes were at distances greater than about 0 . 5 cm from the endocardial wall activity . The effect was observed in the basket phase map when considering the whole atrial tissue ( Fig 8A ) and on the endocardium-blood interface and the basket phase maps when considering the tissue encompassing the CT rotor and its RWE ( Fig 8B ) . But if distance keeps increasing or source shrinking , the influence of far-field sources is not enough to generate IMPSs , as demonstrated by considering only the tissue encompassing the CT rotor ( Fig 8C ) . We hypothesized that reducing the inter-electrode distance , would reduce the percentage of FIPSs . Therefore , we analyzed the effect of modifying the number of electrodes in the basket by decreasing it to 4×6 or increasing it to 16×16 ( Fig 9A ) . Fig 9B–9D are snapshots of the basket phase maps for each of the three positions ( see also S1 Table ) . Increasing the electrodes density at the SVC position improved the ability to detect the CT rotor from 85% ( 4×6 ) to 94% ( 8×8 ) and 97% ( 16×16 ) of the time; at the CT position from 46% ( 4×6 ) to 90% ( 8×8 ) and 94% ( 16×16 ) ; and at the CS position from 35% ( 4×6 and 8×8 ) to 63% ( 16×16 ) . Clearly , improvement was less in the CS position because the basket was located farthest from the CT rotor meandering area . Notably , at an electrode density of 4×6 in either the SVC or CS position , the detection of the RWE was impaired by the appearance of the FIPSs . Increasing the electrode density from 8×8 to 16×16 , improved detection slightly ( 7 to 9% ) for the SVC position , but remained unchanged for the CS position ( 61% ) . Whereas in the SVC and CT positions IMPSs were present 100% of the simulation time regardless of the electrode density , the percentage of time during which FIPSs were present went down considerably ( from 70% to 57% and 0% for the SVC position and from 59% to 13% and 8% for the CT position ) when increasing the electrode density . In the CS position , in addition to a poor detection of the CT rotor , the 4×6 electrode basket yielded an extremely high percentage of FIPSs and IMPSs during the entire simulation time and it was almost impossible to differentiate between them . Increasing the density of electrodes allowed differentiating IMPSs from FIPSs , so the FIPSs percentage was reduced to 40% for 8×8 electrodes and eliminated for the 16×16 electrode basket density . In addition , we computed the phase maps with ~900 interpolated points for the 8×8 basket , and as expected , the number of FIPSs was lower than for the phase maps with 57600 interpolated points . This fact confirmed our hypothesis regarding the effect of increasing the resolution of the phase mapping through increasing the number of interpolated points , which yields a higher number of FIPSs . This effect was also shown when comparing the 8×8 basket with the 16×16 basket: both had 57600 interpolated points but in case of the 8x8 basket the number of interpolated points doubled and the number of FIPSs was higher . To sum up , CT rotor detection accuracy improved slightly when increasing the number of electrodes above 8×8 for the SVC and CT positions . Improvement was significant for the CS position , in which the rotor coverage was poor . In addition , no FIPSs appeared for the SVC and CS positions , whereas the percentage of FIPSs decreased slightly for the CT position . However , accuracy worsened considerably and the percentage of FIPSs increased greatly when decreasing the number of electrodes . Finally , the percentage of IMPSs remained stable regardless of the electrode density . To date the accuracy of mapping AF to localize rotors using panoramic basket catheters has not been validated , in part because in clinical practice the fibrillatory activation patterns are not known . We have used computer simulations to analyze in detail factors affecting AF mapping and localization of rotors . Our results show that rotors may be identified by phase maps of electrical recordings directly from the endocardial surface-blood interface , but less reliable so by phase maps built from basket catheter recordings . Importantly , a potential inaccuracy of the basket maps includes phantom rotors , which may confound targeting of ablation to terminate AF . Our analysis suggests that the appearance of the phantom rotors can be attributed to at least three factors: The distance between the basket electrodes and the endocardial wall ( positioning ) , the distance between the atrial waves and the electrodes ( far-field ) and the inter-electrodes distance ( interpolation ) of data used to create the maps . Therefore , our results suggest that while phase maps based on basket catheters are a powerful tool to map AF and to localize real rotors and other ablative targets , they can also mislead physicians to ablate atrial regions that are in fact free of rotor sources of AF . The distance between the endocardial wall and the basket electrodes depends on the basket´s position within the atria . At each electrode location distance changes non-uniformly , which clearly affects rotor detection . Our results reveal that in addition to the amount of RA area coverage with small distance ( d≤0 . 5cm ) , coverage of the rotor meandering area is important . In our simulations , the basket at the CS position was the farthest from the rotor meandering area and therefore it detected the rotor only 30% of the time , which was much lower than the SVC and CT positions at which the basket was closer to the rotor area and detection was over 90% of the simulation time . On the other hand , the false rotors ( IMPSs and FIPSs ) tend to appear in basket regions with d>0 . 5 cm . Thus , basket positioning with gaps between the electrodes and the endocardium may lead to poor rotor detection and probably low rates of AF termination[18 , 19 , 61 , 62] . Our data agree with results by Narayan et al[15] , in which ablation of drivers slowed but did not terminated AF when the atrial coverage was poor because of the limited size of commercial baskets compared to the large size of the atrium . Narayan et al attributed the failure to the existence of residual sources in the unmapped regions . However , according to our results , unsuccessful AF termination may have been also due to ablation of phantom rotors appearing at electrodes near and distant from the true rotors locus . Unfortunately , according to our simulations , false detection of rotors cannot be excluded , because even short electrode-to-endocardial wall distances do not guarantee the elimination of false rotors . We found that the endocardial-blood interface maps could show false rotors as well , likely because of far-field contribution of sources and the high sensitivity of the phase mapping to low amplitude signals ( Fig 8 ) . Interestingly , when the distance between the basket electrodes and the endocardium increases ( after excluding part of the atrial tissue in the computations ) , IMPSs tend to disappear ( Fig 8C ) , probably because scroll wave filaments originating at endocardial IMPSs do not reach deep into the cavity[52] . Our results confirm that an 8×8-pole mapping basket catheter can yield sufficient spatial resolution for rotor detection when it is properly located in contact with the tissue at the rotor meandering area ( SVC and CT positions in our simulations ) . Increasing the electrode density did not significantly improve rotor detection . We however predict that decreasing the electrode density ( e . g . , 4×6 ) from an optimal level will reduce the ability to detect rotors , while increasing it ( e . g . , 16×16 ) will not substantially alter results if the basket is located close to the rotor area ( a substantial improvement was observed only for the basket at the CS position , where the basket did not cover the meandering area of the rotor ) . Other studies are consistent with our observations . Narayan et al[8] showed that irregular inter-electrodes distances do not alter the sequential activation across adjacent electrodes surrounding a rotor . In addition , the study by Rappel and Narayan[16] suggested that the spatial resolution of a 64-pole mapping basket catheter is adequate to detect rotors , although noise in the EGMs and electrode position might affect accuracy . This is in accordance with our results showing that CT rotor detection was good for a 64-pole basket positioned in the SVC and CT , whereas it was poor for the CS position . Recently Roney et al[63] found that basket catheters are prone to false detections and may incorrectly reveal rotors that are not present , and also that increasing the number of splines up to 16 reduces both the number of false PSs and the number of missing PSs . In general , our results are in accordance with their results . When we increased the number of splines up to 16 , for the three basket positions the false PSs due to the interpolation ( FIPSs ) were strongly reduced and the sensitivity for detection of the real rotor increased ( see S1 Table in the supplemental material ) . However , our study highlights the fact that rotor tracking is more effective if the basket catheter is placed appropriately inside the atrial cavity to ensure extensive coverage of the rotor meandering area . Indeed , false rotors appearing as a result of a larger than critical electrode-to-tissue distance ( i . e . , IMPSs ) will persist even after improving the spatial resolution ( Fig 9 ) . Furthermore , for certain positions of the basket , the rotor would not be detected if it drifts to a poorly covered region , as seen for the SVC position when the rotor migrated toward the IVC and for the CT position when the rotor migrated toward the SVC ( see Fig 4 ) . It should be noted that , in the clinic , if the basket is not large enough , it would not be possible to determine if it is properly located inside a cavity because one would not know a priori the rotors’ location . Averaging all three positions analyzed , detection of the CT rotor with the 8×8 basket occurred 73% of the time , whereas with the 16×16 basket it was detected 85% of the time , which suggests some improvement with the added electrodes . IMPSs also seem to be detected by the basket at all densities , mostly in electrodes separated from the endocardium by >0 . 5 cm . Averaging all three basket positions , IMPSs are detected 77% and 81% of the time with the 8×8 and 16×16 basket , respectively . On the other hand , false rotors due to interpolation ( i . e . , FIPSs ) appear only when inter-electrode distances are large ( reducing the inter-electrode distance by increasing the density to 16×16 substantially decreased , or eliminated , the occurrence of FIPSs ) . Averaging all three positions , FIPSs are detected 37% and 4% of the time with the 8×8 and 16×16 basket , respectively . Overall , our simulations suggest that the probability of ablating an erroneous target would be higher than the probability of correctly ablating a target when using a small 8×8 basket to guide ablation . However , the probability of correctly ablating a target would increase substantially when using a small 16×16 basket . We need to consider several potential limitations of our study . The 3D atrial model is anatomically and electrophysiologically realistic , but is simplistic regarding wall thickness and ionic heterogeneous details . Although it does not alter our main conclusion , the preferable distance to avoid imaginary rotors ( IMPSs; < 0 . 5 cm in our simulations ) could be dependent on the anatomy of the atrial model; for example , it is likely that if the atrial wall thickness would change , this distance would also change . In addition , we have presented simulations for a single scenario of a relatively large rotor area without additional wavebreaks . Such considerations limit our ability to extrapolate quantitatively the results to other fibrillatory wave propagation scenarios . In addition , our study used a single signal processing protocol that included band-pass filtering and the Hilbert Transform , together with an automatic PS detection . We did not explore other EGM processing methods that could have affected the rotors detection , however the protocol used is considered generic to phase mapping and as such very clinically relevant . Another factor limiting the accuracy of basket-based phase maps is the quality of the signals used here compared to the actual clinical EGM signals , which are usually contaminated by far-field effects from the ventricles and noise . Patterns of real atrial waves and signals during AF are probably more complex and with a higher number of artefacts than those simulated here , which could decrease the reliability of the computed phase maps . Finally , we simulated the EGMs recorded by spherical mapping basket catheter located inside the RA and no deformations were applied . Although somewhat unrealistic , this geometrical configuration is well suited to highlight the clinically important effect of varying the distance between the electrodes and the endocardium[15 , 18 , 19] as well as the difference between the basket phase maps and the endocardium-blood interface phase maps . Moreover , our characterization of the various effects of specific basket positions on detection of true and false rotors is based on solid theoretical principles that are commonly accepted in cardiac electrophysiology and provide insights into the drawbacks of using basket catheter-based phase mapping of AF rotor sources . We demonstrate that atrial rotor detection can be achieved by a phase analysis of multi-electrode basket catheter positioned at any distance from the atrial tissue , but preferably placed closer than 0 . 5 cm to the atrial tissue as basket electrodes far from the tissue tend to produce false rotors in addition to real rotors due to the increased effect of distant activity . We further demonstrate that distant activity can also produce imaginary PSs in phase maps even at short electrode-to-endocardial wall distances and without interpolation . Overall , maintaining the basket electrode grid at 8×8 or higher seems sensitive enough for detection of large area rotors , although accuracy will vary depending on the position of the basket inside the atrial cavity and the number of electrodes . Importantly , although basket catheters are currently used to guide patient-specific ablation of the AF drivers , spurious targets in the form of phantom rotors cannot be excluded and all detected rotors should be cautiously considered . | In computer simulations we determined the accuracy of using multiple-electrode basket catheters to detect atrial fibrillation sources ( rotors ) during an ablation procedure . We used a realistic 3D atrial model and a virtual multiple-electrode basket catheter placed in three different positions inside the right atrium . The ability to detect a true rotor depended on three major factors: 1 ) the position of the basket inside the atrium , 2 ) the distance between the electrodes and the atrial wall , and 3 ) the inter-electrode distance . When the electrodes were not in full contact with the atrial wall far-field sources predominated on the recorded signal . As a consequence , phantom rotors were generated on the maps . Interpolation of the signals increased the false rotor incidence , whereas increasing the electrode density decreased it . We conclude that the appearance of false rotors might contribute to AF ablation procedure failure because the physician may erroneously target atrial tissue locations where no true rotor exists . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results"
] | [
"biotechnology",
"medicine",
"and",
"health",
"sciences",
"engineering",
"and",
"technology",
"cardiovascular",
"anatomy",
"catheters",
"membrane",
"potential",
"electrophysiology",
"endocardium",
"veins",
"interpolation",
"mathematics",
"cardiology",
"cardiac",
"atria",
"a... | 2018 | Factors affecting basket catheter detection of real and phantom rotors in the atria: A computational study |
Legionella pneumophila , the etiological agent of Legionnaires’ disease , replicates intracellularly in protozoan and human hosts . Successful colonization and replication of this pathogen in host cells requires the Dot/Icm type IVB secretion system , which translocates approximately 300 effector proteins into the host cell to modulate various cellular processes . In this study , we identified RavK as a Dot/Icm substrate that targets the host cytoskeleton and reduces actin filament abundance in mammalian cells upon ectopic expression . RavK harbors an H95EXXH99 motif associated with diverse metalloproteases , which is essential for the inhibition of yeast growth and for the induction of cell rounding in HEK293T cells . We demonstrate that the actin protein itself is the cellular target of RavK and that this effector cleaves actin at a site between residues Thr351 and Phe352 . Importantly , RavK-mediated actin cleavage also occurs during L . pneumophila infection . Cleavage by RavK abolishes the ability of actin to form polymers . Furthermore , an F352A mutation renders actin resistant to RavK-mediated cleavage; expression of the mutant in mammalian cells suppresses the cell rounding phenotype caused by RavK , further establishing that actin is the physiological substrate of RavK . Thus , L . pneumophila exploits components of the host cytoskeleton by multiple effectors with distinct mechanisms , highlighting the importance of modulating cellular processes governed by the actin cytoskeleton in the intracellular life cycle of this pathogen .
Legionella pneumophila is a ubiquitous Gram-negative bacterium that lives as a parasite of fresh water amoebae in the environment . It is also an important pathogen for humans; inhalation of L . pneumophila-contaminated aerosols by immune-compromised individuals can lead to a severe form of pneumonia , which is referred to as Legionnaires’ disease [1] . It is believed that protozoans hosts provide the major evolutionary pressure for L . pneumophila to acquire and maintain virulence factors essential for its intracellular survival and replication in human macrophages [2] . One hallmark of L . pneumophila infection is the formation of an ER-derived membrane-bounded vacuole known as the Legionella-containing vacuole ( LCV ) , which bypasses the default endocytic pathway that ultimately delivers phagocytosed particles to the lysosome . The biogenesis and development of the LCV strictly requires the Dot/Icm type IV secretion system [3 , 4] , through which approximately 300 protein substrates are translocated into the host cytosol . These proteins , also called effectors , function to modulate a wide spectrum of host cellular pathways , including membrane trafficking , ubiquitination , autophagy , immune responses , and the actin cytoskeleton [5–13] . Despite intensive efforts , only a small proportion ( about 10% ) of the ~300 Dot/Icm effector proteins have been characterized biochemically [14 , 15] . The 42-kDa actin protein assembles into filaments within cells to construct a pervasive and dynamic cytoskeleton , which plays a crucial role in diverse cellular processes including cell migration , cytokinesis , endocytosis and vesicle trafficking [16] . Therefore , it is not surprising that many pathogens have evolved effective strategies to target actin and/or proteins involved in the regulation of actin activity . Intracellular bacterial pathogens such as species of Listeria , Shigella , Rickettsia and Burkholderia take advantage of distinct host actin polymerization machineries to facilitate their movement within the host cytosol and/or their cell-to-cell spread [17] . Salmonella enterica Typhimurium modulates the actin cytoskeleton to gain entry into non-phagocytic cells [18] . Chlamydia trachomatis coopts the function of actin filaments and intermediate filaments to stabilize its replicative vacuole in epithelial cells [19] . Apart from these , bacterial proteins directly modifying actin monomers have also been identified . The best-studied modification is ADP-ribosylation of actin by the C2 toxin from Clostridium botulinum , which modifies Arg-177 of actin , leading to the inhibition of actin polymerization [20] . In contrast , the Photorhabdus luminescens Tc toxin ADP-ribosylates the Thr-148 residue to promote actin polymerization , facilitating the formation of actin aggregates [21] . Bacterial proteins that cleave actin have also been identified; the metalloprotease ECP32 from Serratia proteamaculans cleaves actin , and ectopic expression of this protein enables nonpathogenic E . coli to invade eukaryotic cells [22] . Targeting host actin cytoskeleton by L . pneumophila virulence factors has emerged as an exciting area of research . At least three Legionella Dot/Icm substrates have been shown to modulate distinct cell biological aspects of actin cytoskeleton components . VipA is an actin nucleator , which localizes to actin patches and endosomes during infection and promotes actin polymerization [13]; Ceg14 co-sediments with filamentous actin and inhibits actin polymerization by an unknown mechanism [12]; LegK2 is a kinase that phosphorylates ArpC1b and Arp3 , two subunits of the Arp2/3 complex , thus inhibiting actin polymerization on the LCV [11] . Considering the importance of the actin cytoskeleton in cellular processes and extensive functional redundancy among Legionella effectors , we hypothesized that more Dot/Icm effectors function to target the actin cytoskeleton . In a screening for Dot/Icm substrates capable of modulating the actin cytoskeleton , we identified RavK as an effector that disrupts the actin cytoskeleton of mammalian cells . We further provide evidence that RavK is a zinc-dependent metalloprotease that specifically cleaves actin and abolishes its polymerization activity . Together with earlier reports on VipA , LegK2 and Ceg14 , our results add to a growing body of evidence that L . pneumophila utilizes multiple proteins to modulate different aspects of the host actin cytoskeleton in its intracellular life cycle .
To identify effectors that target the actin cytoskeleton , we screened a GFP fusion library of Dot/Icm substrates [23] for their ability to alter the morphology of the mammalian actin cytoskeleton . Given the essential role of the actin cytoskeleton in cell viability , disruption of its structure most likely is detrimental; we thus began our screening by examining the effects of Dot/Icm substrates known to be toxic to yeast [12 , 24–26] . From the first eight candidates screened , we found that ectopic expression of effector RavK ( Lpg0969 ) led to the abolishment of the actin cytoskeleton in COS-1 cells ( Fig 1 ) . Interestingly , overexpression of Lpg0944 caused a detectable rearrangement of actin cytoskeleton with more F-actin accumulating on the plasma membrane ( Fig 1 ) . In contrast , cells transfected to express the other 6 effectors showed only very minor or undetectable changes in the structure of the actin cytoskeleton compared with those expressing GFP ( Fig 1 , S1 Fig ) . The strong and clear phenotype associated with RavK prompted us to further investigate its mechanism of action . RavK was originally identified in a screening for L . pneumophila Dot/Icm substrates by its ability to restore the translocation of the transfer-deficient mutant SidCΔC100 , and therefore was designated as RavK ( region allowing vacuole co-localization K ) [27] . Dot/Icm-dependent translocation of RavK was independently demonstrated using the CCF4/β-lactamase reporter assay [5] . Probably due to the need for effectors that effectively thwart the host defense in the initial phase of infection , the expression of many Dot/Icm substrates is induced during the post-exponential phase , when L . pneumophila concomitantly enters the transmissive phase and becomes primed for a new round of infection [24] . We therefore examined the level of RavK at different time points throughout the growth cycle of L . pneumophila in broth . Interestingly , the expression of RavK was highly induced in exponentially growing bacteria ( 6–18 h ) ( OD600 between 0 . 4 and 3 . 0 ) ; the protein was barely detectable in the lag ( 0–6 h ) or the post-exponential phase ( 18–24 h ) ( Fig 2 ) , which indicates that RavK likely plays a role in the replicative phase during L . pneumophila infection . To understand the mechanism of action of RavK , we first performed sequence analysis of the protein to search for the presence of motifs suggestive of known biochemical activity . We manually scanned the sequence of RavK against the “PROSITE collection of motifs” [28] , and found that RavK harbors an H95EXXH99 motif present in diverse metalloproteases [29] ( Fig 3A ) . To determine the role of this motif in the activity of RavK , we introduced mutations in H95 , E96 and H99 , respectively . Next we assessed the effects of these mutations on the activity of RavK by examining their toxicity to yeast; while not affecting the stability of RavK , each of these mutations completely abolished the toxicity to yeast ( Fig 3B and 3C ) . To examine whether the H95EXXH99 motif is required for the disruption of actin cytoskeleton , we expressed GFP-RavK , GFP-RavKH95A or GFP in COS-1 cells and labeled the actin cytoskeleton with Texas-red-conjugated phalloidin . Relative F-actin levels in transfected cells were analyzed by calculating the integrated pixel density of phalloidin fluorescence of outlined individual cell . Our results indicate that the total F-actin levels in GFP-RavK-expressing cells were significantly lower than those in cells expressing GFP or GFP-RavKH95A ( Fig 3D and 3E ) . We also found that cells expressing GFP-RavK were significantly smaller than that those expressing GFP or GFP-RavKH95A ( Fig 3F ) , indicating that ectopic expression of RavK caused shrinkage in COS-1 cells . In comparison to the RavK-mediated morphological alterations in COS-1 cells , ectopic expression of GFP-RavK caused a clear cell-rounding phenotype in HEK293T cells . Consistently , the observed phenotype in HEK293T cells also depends on the H95EXXH99 motif ( S2A Fig ) . The different responses to RavK by COS-1 and HEK293T cells may be due to the expression level , variations in the cellular level of the protein targeted by RavK , or a combination of both . Nevertheless , these results suggest that RavK is a metalloprotease that potentially target components of the host cytoskeleton . Actin exists in cells as both free monomer , called G-actin ( globular actin ) , and as polymeric microfilaments , called F-actin ( filamentous actin ) [30] . The RavK-induced reduction of the phalloidin-stainable F-actin in COS-1 cells can be accounted for by at least two possibilities . First , RavK directly reduces the total pool of actin within the cells by mechanisms such as proteolytic cleavage . Second , RavK somehow tilts the balance toward G-actin and reduces the pool of F-actin . To distinguish between these two possibilities , we compared the total actin level between cells expressing RavK and the RavKH95A mutant and found that cells expressing wild-type RavK contained much lower levels of total actin than that of cells expressing RavKH95A or GFP ( Fig 4A and 4B ) , indicating that RavK reduces the abundance of total actin in COS-1 cells . Similarly , RavK expression also reduced total actin level in HEK293T cells ( S3A Fig ) . The reduction of cellular actin levels by RavK can be caused by directly degrading the protein or by initiating a signaling cascade that leads to lower cellular actin levels . A direct approach to distinguish between these two models is by incubating recombinant RavK with total lysates of mammalian cells and examining the levels of actin . To obtain active RavK protein for such biochemical assays , we made numerous attempts to express epitope-tagged RavK for affinity purification from E . coli , none of the used tags such as His6 , His6-Sumo , and GST allowed us to obtain soluble full-length RavK ( S4A Fig ) . We therefore initiated a screening to identify truncated alleles of RavK that would potentially be soluble and functional for biochemical studies . A series of RavK deletion mutants were constructed by removing residues from its C-terminal end ( the H95EXXH99 motif localizes toward its N-terminal portion ) . Whereas deletion of 50 residues from the C-terminal end led to a mutant that retained the toxicity to yeast , a mutant lacking 100 amino acids from the same end abolished its toxicity ( S4B and S4C Fig ) . Consistent with its toxicity to yeast , RavKΔC50 still caused cell rounding in HEK293T cells ( S4D and S4E Fig ) . Notably , the ΔC50 deletion greatly increased the solubility of RavK , which allowed us to obtain sufficient recombinant protein for biochemical experiments ( S4A Fig ) . To determine the activity of the recombinant RavKΔC50 , we incubated lysates of COS-1 cells with His6-RavKΔC50 or His6-RavKΔC50H95A at 22°C for 1 h . Wild type RavKΔC50 but not the H95A mutant caused a reduction of full-length actin and produced an actin fragment clearly smaller than the original protein in the cell lysates . Furthermore , the reduction of actin can be inhibited by the metal ion chelator EDTA ( Fig 4C ) . Similar results were observed with lysates of HEK293T cells ( S3C Fig ) . Thus , RavK is a metalloprotease , which is able to cleave actin in lysates of mammalian cells in an H95ExxH99 motif-dependent manner . We next tested whether any host factor is required for the cleavage of actin by RavK by mixing human non-muscle actin ( 85% β-actin and 15% γ-actin ) with His6-RavKΔC50 or His6-RavKΔC50H95A for various time durations . Incubation with His6-RavKΔC50 but not with the H95A mutant produced a smaller actin fragment ( Fig 4D ) and the size difference between these two fragments is similar to that observed in experiments using total cell lysates . Consistent with earlier observations , the activity of RavK is sensitive to EDTA . Thus , RavK is a metalloprotease that cleaves actin without the requirement of any other host proteins . Among the three major groups of actin ( α , β , γ ) identified in vertebrates , the α-actin is the major constituent of the contractile apparatus in muscle cells , whereas the β and γ actin coexist in most of non-muscle cells as a component of cytoskeleton [31] . We thus tested whether RavK has a preference toward specific actin isoforms . Since the protein sequence of commercially available rabbit skeletal muscle actin is identical to that of human skeletal muscle actin , we used rabbit muscle actin in this assay for comparison to human non-muscle actin . The same amount of rabbit muscle actin and human non-muscle actin was treated with equal amount of His6-RavKΔC50 . As early as 2 min post treatment , significantly more cleaved product was detected in reactions with non-muscle actin than those with muscle actin . When the reaction was allowed to proceed for 128 min , more than 80% of non-muscle actin was cleaved , whereas only approximately 25% of muscle actin that was cleaved in this experimental duration ( Fig 4E and 4F ) . Thus , RavK cleaves the non-muscle actin more efficiently than the muscle actin . As RavK harbors an H95EXXH99 motif that is common for zinc binding , we further tested whether zinc is required for the activity of RavK . Seven different metal ions including zinc were tested for their ability to restore the activity of metal ion-free RavK . Indeed , Zn2+ was able to restore the activity of RavK ( S5 Fig ) . Notably , whereas such divalent ions as Co2+ , Ni2+ , Cu2+ , Ca2+ or Mg2+ cannot detectably restore the activity of RavK , Mn2+ was able to restore the activity of RavK at levels comparable to those of Zn2+ , probably due to their similarity in chemical properties [32] . To investigate whether the RavK-mediated cleavage of actin occurs during bacterial infection , we infected mammalian cells expressing 4xFlag-tagged actin with wild-type L . pneumophila or its derivatives . Cleaved actin was only detected in samples infected with the ΔravK strain expressing RavK from a multi-copy plasmid but not in samples infected by the ΔravK strain or ΔravK overexpressing RavKH95A , indicating that actin is cleaved during L . pneumophila infection in a RavK-dependent and more specifically metalloprotease motif-dependent manner ( Fig 5A ) . The cleaved form of actin was not observed in samples infected by wild-type L . pneumophila , which may be attributed to less translocated RavK in host cytosol compared to samples infected by ΔravK overexpressing RavK . Considering the expression level of RavK is much higher in ΔravK overexpressing RavK than in wild-type L . pneumophila ( Fig 5B ) , it is almost certain that more RavK was translocated to host cells by the overexpressing strain , which caused detectable cleavage of Flag-actin . Yet , in both cases the amount of translocated RavK was below the detection capacity of immunoblotting with our RavK-specific antibody ( Fig 5C ) . Despite multiple attempts using different infection conditions such as variations in multiplicity of infection ( MOI ) , infection time and host cells , we were unable to detect the cleavage of endogenous actin even in infections using the strain overexpressing RavK ( Fig 5D ) . The inability to detect the reduction of actin or the cleaved product may attribute to low stability of the cleaved product in cells , the quality of the antibodies used for detection or the potential compensatory effects from the hosts , or a combination of these factors . To determine the cleavage site of actin by RavK , we incubated non-muscle actin with His6-RavKΔC50 at 22°C for 1 h . Samples resolved by SDS-PAGE were detected by Coomassie brilliant blue staining . Protein bands corresponding to both the uncleaved , full-length and the cleaved products were excised , digested with trypsin and sequenced by mass spectrometry ( S6A Fig ) . Analysis of the detected tryptic fragments revealed that the semi-tryptic peptide -Y337SVWIGGSILASLST351- was present in the cleaved protein but not in the full length protein , suggesting that the cleavage site lies between Thr351 and Phe352 ( Fig 6A ) . Consistent with this notion , the abundance of the N-terminal peptide -D2DDIAALVVDNGSGMCK18- was similar between these two proteins , whereas the abundance of the C-terminal peptide -Q360EYDESGPSIVHR372- was significantly higher in the full-length protein than in the cleaved product ( Fig 6A ) , further suggesting that the cleavage site identified by this method is reliable . We confirmed the identified cleavage site by mutating Phe352 into an Ala in β-actin . A Flag-tagged β-actinF352A gene was expressed in HEK293T cells by transfection . Immunoprecipitated Flag-β-actinF352A eluted with the Flag peptide was incubated with His6-RavKΔC50 and the cleavage product was detected by immunoblotting with the Flag-specific antibody . RavK treatment of similarly purified wild type Flag-β-actin yielded two protein bands with a molecular weight difference resemblying that observed in experiments with purified actin ( Fig 6B ) . In contrast , only one single protein corresponding to the size of uncleaved actin was detected in samples expressing the actinF352A mutant ( Fig 6B ) , establishing that Phe352 is important for RavK-mediated cleavage . Residues around the cleavage site often provide the strucutral context important for recognition by proteases [33]; we therefore constructed a series of mutants with substitution mutations in sites adjacent to Phe352 and examined their sensitivity to RavK . Our results indicate that Leu349 , Ser350 , Thr351 are indispensible for RavK-mediated cleavage , whereas Gln353 and Gln354 are not essential ( Fig 6C ) . With the exception of Ser350 , which is replaced by a Thr residue in yeast actin , all the other residues tested in our mutational analysis are conserved among all human actin isoforms and yeast actin ( Act1 ) ( S6B Fig ) . Thus , it is likely that the yeast toxicity of RavK is due to its cleavage of yeast actin . We tested this hypothesis by incubating Flag-Act1 with RavKΔC50 . Incubation of wild-type RavK but not the RavKH95A resulted in the production of a smaller Act1 fragment , and the fragment was absent in the reaction receiving EDTA ( Fig 6D ) , indicating that RavK cleaves Act1 in a metalloprotease activity-dependent manner . Thus , the protease activity against actin attributes to the cytotoxicity of RavK in yeast . Actin is one of the most abundant proteins in eukaryotic cells , it is possible that the cleavage by RavK we observed is due to non-specific activity . To test whether actin is a bona fide target of RavK , we set out to examine whether overexpressing a cleavage-resistant actin variant in HEK293T cells could rescue the cell rouding phenotype mediated by RavK . Overexpression of the actinF352A mutant did not cause any disernable effects in mammalian cells , suggesting that this mutation did not overtly affect the function of actin . If actin is the true substrate of RavK , cells overexpressing actinF352A should become resistant to damage caused by the protease . Overexpression of wild-type actin did not reduce the percentage of rounded cells induced by RavK , which was similar to samples receiving only the construct for RavK ( Fig 7 ) . In contrast , overexpression of actinF352A in HEK293T cells almost completely abrogated the cell rounding phenotype , although these cells expressed RavK at levels similar to other samples ( Fig 7 ) . The ability of actinF352A to effectively suppress the RavK-induced phenotypes further establishes that actin is a bona fide cellular target of RavK . Actin exists as both G-actin and F-actin in the cell and the transition between these two forms in response to cellular needs is precisely regulated [30] . Given the importance of actin polymerization in its function , we tested whether the RavK-cleaved actin retains the ability to form actin filaments . The formation of filaments by G-actin can occur spontaneously under certain conditions , which can be measured by sedimentation after high-speed centrifugation [34] . We therefore determined the polymerization activity of actin after RavK-mediated cleavage . As complete cleavage of non-muscle actin by RavK cannot be achieved even after extended incubation , a mixture consisting of cleaved and uncleaved actin was used in this and following assays . Non-muscle actin that had been incubated with RavKΔC50 or RavKH95AΔC50 at 22°C for 2 h was induced to polymerize for 60 min . The formation of actin filaments was determined by its presence in pellets after ultracentrifugation . Similar to mock-treated actin , in reactions containing actin that had been incubated with RavKH95AΔC50 , the majority of actin was in the pellets , indicative of robust polymerization ( Fig 8A and 8B ) . In contrast , in reactions containing RavKΔC50 , only approximately 40% of the actin was present in the pellet , indicating the cleaved product is defective in polymerization . To further validate our findings based on the high-speed centrifugation assay , we examined the ability of actin1-352 to form actin filaments by a pyrene-labeled actin nucleation assay . Two-hour treatment with RavKH95AΔC50 only slightly decreased the polymerization property of non-muscle actin compared with the control samples receiving no additional protein . On the other hand , incubation with RavKΔC50 for the same time duration drastically reduced the ability of actin to form polymers for non-muscle actin ( Fig 8C and 8D ) , further confirming RavK-cleaved actin is defective in forming actin filaments . RavK is present in 23 out of 41 sequenced Legionella species , and is one of the 74 effectors which are shared by more than 20 different Legionella species [35] . Such a high prevalence suggests an important role of this protein in the interactions of the bacteria with their hosts . To study the role of RavK during L . pneumophila infection , we constructed an in-frame deletion mutant of this gene and investigated the intracellular replication of this mutant in primary mouse macrophages and the protozoan host Dictyostelium discoideum . Similar to many Dot/Icm substrates , the results show that the absence of RavK does not affect the uptake by host cells or the intracellular replication capacity of L . pneumophila in mouse macrophages or D . discoideum ( S7A and S7B Fig ) . It has been reported that two other Legionella effectors Ceg14 and VipA also target host actin cytoskeleton . Ceg14 co-sediments with filamentous actin and inhibits actin polymerization [12] , whereas VipA is an actin nucleator , promoting actin polymerization [13] . Since both Ceg14 and RavK negatively affect the actin polymerization , we investigated whether the absence of both of them would cause any intracellular growth defects . The intracellular growth of ceg14/ravK double knockout strain was determined and the results showed that it still grows as proficiently as the wild-type strain in either mouse macrophages or D . discoideum ( S7A and S7B Fig ) . Recently , another Legionella effector LegK2 was shown to inhibit actin polymerization by phosphorylating the Arp2/3 complex [11] . We therefore made a ceg14/ravK/legk2 triple knockout mutant to examine the potential functional redundancy of these three proteins . The ceg14/ravK/legK2 triple mutant manifests ~10 fold growth defect compared to wild-type strain in D . discoideum . However , this defect is comparable to that of the legK2 mutant ( S7C Fig ) , indicating the absence of RavK and Ceg14 does not confer any further growth defects to the legK2 mutant . Furthermore , the intracellular growth defect of the triple mutant can be complemented by legK2 but not ravK or ceg14 ( S7D Fig ) .
Actin cytoskeleton is a common target exploited by many bacterial pathogens , both intra- and extracellular . Generally , bacterial effectors identified to date modulate host actin cytoskeleton by two different mechanisms of action . First , many bacterial effectors interfere with endogenous actin regulation pathways . Examples in the group include effectors that target the small GTPases Rho , Rac and Cdc42 , master regulators of the actin cytoskeleton , by either distinct post-translational modifications [36–40] , or the regulation of their GTP binding status [41–43] . The second mechanism of action is to directly modify the actin molecule by means of posttranslational modifications such as ADP-ribosylation , or by crosslinking or proteolysis . ADP-ribosylation of actin leads to either promotion or inhibition of actin polymerization depending on the residues being modified . ADP-ribosylation of Arg-177 by the C2 toxin from Clostridium botulinum inhibits actin polymerization [20] , in contrast , the same modification of Thr-148 by the Tc toxin from Photorhabdus luminescens promotes actin polymerization [21] . Actin cross-linking proteins secreted by Vibrio and Aeromonas species induce the production of actin oligomers that strongly inhibit Formin-mediated actin polymerization [44] . Proteolysis of actin by bacterial proteins has also been documented; the metalloprotease ECP32 from Serratia grimesii cleaves actin , and ectopic expression of this protein enables nonpathogenic E . coli to invade eukaryotic cells [22] . In this study , we have shown that the Legionella Dot/Icm substrate RavK is a zinc-dependent metalloprotease that specifically cleaves actin to disrupt the actin cytoskeleton of host cells . Unlike ECP32 from Serratia , RavK does not affect the uptake of bacteria ( S7A and S7B Fig ) . Instead , RavK is likely to play a role in the replicative phase of L . pneumophila during infection , which is supported by the high level expression of ravK at the exponential growth phase in bacteriological medium ( Fig 2A and 2B ) . The activity of RavK toward actin generates products that can be further degraded in the cell , thereby causing the reduction of total actin levels ( Fig 4A–4C ) , which may explain our inability to detect cleaved actin during L . pneumophila infection . RavK is able to cleave purified actin in reactions free of other proteins , indicating that the cleavage of actin by RavK does not require additional proteins from the host or L . pneumophila ( Fig 4D ) . Interestingly , RavK exhibits a preference for non-muscle ( 85% β-actin and 15% γ-actin ) over muscle actin ( α-actin ) . The primary sequences of the three actin isoforms near the cleavage site are identical ( S6B Fig ) , suggesting that the conformation of these actin isoforms at the cleavage site may vary , causing differences in the accessibility for the enzyme and differences in the cleavage efficiency . These results are in line with the fact that natural protozoan hosts of L . pneumophila such as D . discoideum , contain actin which shares a higher level identity of amino acid composition with human β-actin ( 93% ) and γ-actin ( 93% ) than with α-actin ( 89% ) . Actin is a 375-amino acid polypeptide , which folds into two major domains . The two domains are separated by a deep cleft , in which relatively few interactions occur between the two domains . The polypeptide crosses the cleft twice in the middle of the cleft , dividing the cleft into two parts—upper and lower . The upper cleft is responsible for nucleotide binding , whereas the lower cleft is important for the interaction between actin subunits within the actin filaments . The lower cleft is lined by 11 predominantly hydrophobic residues including Leu349 , Thr351[30] . Of note is that RavK cleaves actin at a site between Thr351 and Phec352 , which locates on the outside end of the lower cleft , suggesting that cleavage of actin by RavK may interfere with the interaction between actin subunits in the filament and therefore inhibits the G-actin/F-actin transition . In agreement with this notion , in both the sedimentation and the pyrene-labeled actin nucleation assay , cleaved actin is defective in forming actin filaments ( Fig 8A–8D ) . During L . pneumophila infection , translocated RavK cleaves Flag-tagged actin into a smaller form ( Fig 5A ) , the size difference between the full-length and cleaved actin is similar to that observed in in vitro cleavage assay ( Figs 4D and 5A ) , indicating that RavK likely cleaves actin in the same way under these conditions . Even though we can observe a clear cleaved Flag-tagged actin during the infection of ΔravK strain overexpressing RavK , we were unable to detect a reduction in endogenous actin during infection ( Fig 5D–5E ) . Considering that only a small proportion of Flag-tagged actin is cleaved during infection ( Fig 5A ) , and the fact that actin is very abundant in the cell , it is possible that the reduction of endogenous actin is too minute to be detected by the immunoblotting-based method . It is also possible that the host cell has a compensatory mechanism that once actin cytoskeleton is impaired due to a reduction of actin level , more actin will be synthesized to maintain the integrity of the actin cytoskeleton . At least three Legionella Dot/Icm substrates have been shown to modulate the activity of components of the actin cytoskeleton . VipA is an actin nucleator , which localizes to actin rich regions and endosomes and interferes with the Multivesicular Body ( MVB ) pathway [13]; Ceg14 is a cytosolic protein , which inhibits actin polymerization by an unknown mechanism [12]; whereas LegK2 is a kinase , which localizes to the LCV and phosphorylates two subunits of the Arp2/3 complex to inhibit actin polymerization on the LCV [11] . Our demonstration of RavK as an effector that targets the actin cytoskeleton by cleaving actin indicates that L . pneumophila modulates this important host cellular component by diverse mechanisms . Given the essential role of actin in cellular processes , it is tempting to speculate that RavK targets to specific organelles such as the LCV , where it locally affects the function of actin in concert with effectors such as LegK2 to promote the biogenesis of the LCV . Unfortunately , we were unable to examine this hypothesis by directly staining for RavK during L . pneumophila infection due to low abundance of translocated RavK . Alternatively , translocated RavK may be quickly targeted for degradation by host proteases . Nevertheless , the low abundance of translocated RavK is consistent with its activity against an essential host protein . The distinct effects of multiple effectors on the actin cytoskeleton suggest the necessity of a coordinated modulation of the actin cytoskeleton at different levels . Whether these effectors directly balance the effects conferred by one or the other remains to be determined . Alternatively , given their different localization within the cell , it is plausible that these effectors regulate the actin cytoskeleton in an organelle-specific manner during L . pneumophila infection . The fact that the ravK/ceg14/legK2 triple mutant ( Δ3 ) only has a relatively small growth defect in D . discoideum suggests the presence of additional Dot/Icm substrates targeting actin cytoskeleton , reiterating a significant functional redundancy among Dot/Icm substrates [45] . The lack of a strong defect of the Δ3 mutant in intracellular replication or in the cell biological events associated with L . pneumophila infection [14 , 15] makes it difficult to determine the benefit of targeting the actin cytoskeleton by the pathogen under current experimental conditions . Further studies are warranted to identify additional effectors that target the actin cytoskeleton and to elucidate their mechanisms of action .
All animal use procedures were in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Purdue Animal Care and Use Committee ( PACUC ) ( Protocol number 04–081 ) . Bacteria strains used in this study were listed in S1 Table . E . coli strains were grown in Luria broth ( LB ) medium and was supplemented with antibiotics when necessary . The L . pneumophila strains used in this study were derivatives of the Philadelphia-1 strain Lp02 [46] . L . pneumophila was grown and maintained in CYE medium according to a standard procedure [47] . The ravK , ceg14 and legK2 in-frame deletion mutants were constructed as described [48] . Briefly , for the construction of each knock-out plasmid , two pairs of primers were designed so that the target gene was replaced by 32 amino acids including the first and last 15 residues encoded by the gene and 2 residues encoded by the recognition site of BamHI . For complementation experiments , the gene was expressed on the RSF1010-derived plasmid pZL507 [48] . For expression in mammalian cells , ravK or each of these genes was inserted into pEGFP-C1 ( Clontech ) or pFlag-CMV ( Sigma ) . All the plasmids used in this study are listed in S2 Table and the sequences of all primers are in S3 Table . Yeast strains used in this study were W303 [49] and its derivatives ( S1 Table ) . Yeast strains were grown in yeast extract , peptone , dextrose medium ( YPD ) medium or appropriate amino acid dropout minimal media at 30°C [26] . Yeast transformation was performed with the lithium acetate method [50] . Yeast cell lysates for protein analysis were prepared as described [48] . The ORF of ravK or its derivatives were inserted to pSB157 [51] to generate pGal::ravK ( or pGal::ravK derivatives ) , which were digested with StuI and transformed into yeast strain W303 . To determine the yeast growth arrest induced by RavK , overnight cultures of relevant yeast strains grown in liquid selective medium containing glucose were serially diluted 5-fold , and 8μL of each dilution was spotted onto solid medium containing galactose or glucose . Plates were incubated at 30°C for 48 h before images were acquired . COS-1 , HEK293 and HEK293T cells were obtained from the American Type Culture Collection ( Rockville , MD ) and were cultured in Dulbecco’s modified minimum Eagle’s medium ( DMEM ) supplemented with 10% FBS fetal bovine/calf serum ( FBS ) . Bone marrow-derived macrophages were prepared from A/J mice following the standard protocol [52] . For transient expression of exogenous proteins in HEK293T cells , 5 μL Lipofectamine 2000 ( Invitrogen ) was used to introduce 2 . 5 μg plasmids into mammalian cells per 6-well plate well at a cell confluency of 80% . For transient expression of exogenous proteins in COS-1 cells and HEK293 cells , 5 μL Lipofectamine 3000 and 5 μL P3000 ( Invitrogen ) were used to introduce 2 . 5 μg plasmids into cells per 6-well plate well at a cell confluency of 80% . To purify His6-RavKΔC50 and His6-RavKH95AΔC50 , the appropriate gene fragments were inserted into the pQE30 plasmid ( Qiagen ) to generate pZL1221 and pZL1222 , respectively . For protein production , 20-mL overnight culture of E . coli XL1blue harboring pZL1221 or pZL1222 were diluted into 800 mL LB medium ( 100 μg/mL ampicillin ) and were allowed to grow at 37°C to OD600 = 0 . 6–0 . 8 . After the IPTG was added to a concentration of 0 . 1 mM , the cultures were induced at 18°C for 16–18 h for protein expression . Bacterial cells were collected by centrifugation at 6 , 000g for 5 min and were lysed by sonication in the presence of protease inhibitors and 0 . 2% ( wt/vol ) TritonX-100 . The soluble fractions were collected by centrifugation at 12 , 000g for 20 min and were incubated with Ni-NTA beads at 4°C for 2 h . Proteins bound to Ni2+-NTA beads were washed with 20 mM imidazole and were eluted with 300 mM imidazole . To remove imidazole , eluted proteins were dialyzed twice in 50 mM Tris·HCl ( pH 8 . 0 ) , 50 mM NaCl , 5% ( vol/vol ) glycerol and 1 mM DTT . Polyclonal antibodies against RavK was generated at the Pocono Rabbit Farm and Laboratory using recombinant His6-tagged RavKΔC50 purified from E . coli to immunize a rabbit . The RavK antibody was affinity purified following a standard protocol [53] . The α-actin antibody C4 was purchased from MP Biochemicals ( 0869100 ) and was used at 1:5 , 000 . The α-ICDH , α-GFP , α-PGK , α-Flag , α-tubulin were used as described in an earlier study [48] . Signals from each individual protein were detected by fluorescence dye-conjugated antibodies on an Odyssey detection system ( Li-Cor ) . Rabbit skeletal muscle actin ( >99% pure ) ( Cytoskeleton ) or Human platelet non-muscle actin ( >99% pure ) ( Cytoskeleton ) were incubated with RavKΔC50 or RavKH95AΔC50 in G-actin buffer ( 5 mM Tris-HCl 8 . 0 , 0 . 2 mM CaCl2 , 0 . 2 mM ATP , 0 . 5 mM DTT ) at 22°C for indicated time periods . Protein mixtures were analyzed by SDS-PAGE followed by Coomassie brilliant blue staining . To obtain metal ion-free RavKΔ50 , the protein was treated by 1 mM metal ion chelator 1 , 10-phenanthroline at 25°C for 20 min . After treatment , each of the seven different metal ions was added into an in vitro actin cleavage reaction containing 5 μg actin and 0 . 5 μg metal ion-free RavKΔC50 at a final concentration of 0 . 1 mM . In vitro reactions were allowed to proceed for 2 h before analysis by SDS-PAGE and Coomassie brilliant blue staining . Actin that has been incubated with RavKΔC50 was separated by SDS-PAGE and the bands corresponding to the full-length actin and its cleavage products were excised and digested with trypsin . Peptides were analyzed in an Ekspert nanoLC system 400 ( Eksigent ) coupled to a 5600 TripleTOF mass spectrometer ( AB Sciex ) . Peptides were separated in a capillary C18 column ( 75 μm x 15 cm , ChromXP C18-CL , 3 μm , 120 Å ) with the following gradient: 1 min in 5% solvent B ( Solvent A: 0 . 1% FA and solvent B: 80% ACN/ 0 . 1% FA ) , 5–35% solvent B in 60 min , 35–80% solvent B in 1 min , 6 min in 80% solvent B , 80–5% B in 1 min , and hold in 5% for 11 min . The flow rate was set at 200 nL/min and eluting peptides were directly analyzed in the mass spectrometer . Full-MS spectra were collected in the range of 400 to 2000 m/z and the top 50 most intense parent ions were submitted to fragmentation for 50 milliseconds using rolling-collision energy . Peptides with poor MS/MS spectra were targeted to data-independent acquisition , which enabled collecting high quality spectra . MS/MS spectra searched against the human SwissProt database ( downloaded on July 09 , 2013 ) using Paragon tool of Protein Pilot software ( AB Sciex ) considering biological post-translational modifications and matching peptides were inspected manually . 30-μg G-actin was treated with 3-μg RavKΔC50 , RavKH95AΔC50 or left untreated in G-actin buffer at RT for 2 h . The obtained actin mixtures were precleared by centrifugation at 100 , 000g for 30 min and were used in actin co-sedimentation assays following an established protocol [34] . Briefly , the polymerization was initiated by adding 10× actin polymerization buffer ( 500 mM KCl , 10 mM MgCl2 , 10 mM EGTA , 10 mM ATP ) and was allowed to proceed for 60 min . The samples were subjected to ultracentrifugation at 100 , 000g for 40 min . Supernatants and pellets were analyzed by SDS-PAGE , followed by Coomassie brilliant blue staining . Pyrene-actin polymerization assay was performed as described by Schafer et al with minor modifications [54] . 60-μg non-muscle actin was treated by either His6-RavKΔC50 , His6-RavKH95AΔC50 or left untreated at RT for 2 h . His6-RavKΔC50 , His6-RavKH95AΔC50 were further removed from the cleaved products by passing through a Ni2+-NTA column . The flow-through was collected and precleared by centrifugation at 100 , 000g for 30 min . 2 . 7 μM precleared differentially-treated actin and 0 . 3 μM pyrene-labeled actin were mixed . Upon the addition of 10× polymerization buffer ( 500 mM KCl , 10 mM MgCl2 , 10 mM EGTA , 100 mM imidazole HCl , pH 7 . 0 ) , actin polymerization was monitored by measuring pyrene fluorescence intensity at 1s interval , using a PTI Alphascan spectrofluorimeter ( Photon Technology International , South Brunswick , NJ ) . The excitation and emission wavelengths were set at 365 nm and 407 nm , respectively . Data were collected and processed in Excel ( Microsoft ) and the graph was made with Kaleidograph . 16–18 h after transfection , HEK293T or COS-1 cells were collected and lysed as previously described [48] . Approximately 1mg protein ( in approximately 1 mL ) was used for immunoprecipitation by adding 20 μL agarose beads coated with anti-Flag M2 antibody ( Sigma ) . After incubating at 4°C for 3 h on a rotary shaker , the beads were washed with cold TBS for 4 times and proteins bound to the beads were eluted with 3×Flag peptides following manufacturer’s instructions . The eluted proteins were treated with 1 μg RavKΔC50 or RavKH95AΔC50 in 50 μL G-actin buffer at 22°C for 2 h . The protein mixtures were resolved with SDS-PAGE gel , and detected by the M2 antibody . Immunoprecipitation with yeast lysates was carried out as described [48] . 5×104 COS-1 cells were seeded on 24-well plates and were allowed to grow at 37°C overnight . The next day , plasmids carrying full-length hypothetical L . pneumophila genes were introduced into COS-1 cells by Lipofectamine3000 . 24 h later , cells were fixed in 4% paraformaldehyde in PBS at 25°C for 15 min , permeabilized with 0 . 2% Triton X-100 for 5 min , stained with Texas-red-conjugated phalloidin ( 1:500 ) at 25°C for 30 min , and subjected to imaging analysis . L . pneumophila genes that significantly altered the F-actin patterns of COS-1 cells were subjected to further analysis . Fixed cells stained with Texas-red-conjugated phalloidin were subjected to imaging analysis under an Olympus X-81 fluorescence microscope . Images were acquired from a CoolSNAP HQ2 14-bit CCD camera ( Photometrics ) with identical parameters , and were similarly processed using the IPlab ( BD Biosciences ) and CellSens ( Olympus Life Science ) software package . We quantified the relative F-actin level following an established protocol [55] . Briefly , processed images were imported into ImageJ , and background was subtracted from each image . We then carefully outlined each cell by hand , and measured the integrated pixel density of each cell , which generated the average F-actin content per cell . We also measured the area occupied by each cell , which was shown as the average spread cell area . For infection experiments , L . pneumophila strains were grown to the post-exponential phase as measured by optical density of the culture ( OD600 = 3 . 3–3 . 8 ) and judged by an increase in bacterial motility . For L . pneumophila intracellular growth assay , 4×105 bone marrow-derived mouse macrophages or 5×105 D . discoideum were seeded on 24-well plates and were infected with relevant L . pneumophila strains at an MOI = 0 . 05 at 37°C ( for macrophage ) or MOI = 0 . 1 at 25°C ( for D . discoideum ) . At the indicated time points , cells were treated with 0 . 02% saponin for half an hour and the bacteria number was determined by enumerating colony-forming unit ( CFU ) of appropriately diluted saponin-soluble fractions . HEK293 cells were transfected to express 4×Flag-Actin and FCγRII with pJC119R [56 , 57] for 24 h with Lipofectamine 3000 ( Life Technology ) . Bacteria of relevant L . pneumophila strains were opsonized with rabbit anti-Legionella antibodies [48] at 1:500 for 30 min before infecting the cells at an MOI of 50 for 2 h . Cleared lysates prepared from infected cells were subjected to immunoblotting with M2 antibody ( Sigma ) . For experiments to determine the endogenous actin level during infection , HEK293 cells were transfected to express FCγRII for 24 h with Lipofectamine 3000 ( Life Technology ) , bacterial infections were performed as described above . Cleared lysates prepared from infected cells were probed by immunoblotting with an actin-specific antibody . HEK293 cells were transfected to express FCγRII for 24 h with Lipofectamine 3000 ( Life Technology ) . Bacteria of relevant L . pneumophila strains were opsonized with rabbit anti-Legionella antibodies [48] at 1:500 for 30 min before infecting the cells at an MOI of 50 for 2 h . Infected cells were lysed with 0 . 02% saponin , which lyses membranes of mammalian cells but not of bacterial cells . The lysates were probed for RavK with a specific antibody . Translocation of the effector SidC [58] was probed as a control . Immunoblots were scanned with the Odyssey 3 . 0 ( LI-COR Biosciences ) and quantified with ImageJ . Statistical significance for all relevant data was calculated using the unpaired two-tailed Student t tests , with a p value <0 . 05 being considered as significant difference . | Actin is a core component of the actin cytoskeleton , which plays a crucial role in diverse cellular processes including cell migration , cytokinesis , endocytosis and vesicle trafficking . Therefore , it is not surprising that many pathogens target actin and/or proteins involved in the regulation of actin activity for their benefit . Legionella pneumophila , the etiological agent of Legionnaires’ disease , uses the Dot/Icm type IVB secretion system to transfer effectors into host cells to subvert host cellular processes for its intracellular replication . At least three Dot/Icm substrates , VipA , Ceg14 and LegK2 have been shown to modulate the host actin cytoskeleton . Here , by screening L . pneumophila Dot/Icm substrates that alter the actin cytoskeleton in mammalian cells , we have identified RavK as an additional effector that specifically disrupts actin organization . RavK harbors a canonical metalloprotease motif , which is essential for the RavK-mediated actin cytoskeleton disruption and cell- rounding phenotypes . We further demonstrate that RavK directly cleaves actin , generating a fragment with a diminished capacity to form actin filaments . Our results reveal a new mechanism for which an intravacuolar bacterium disrupts actin cytoskeleton through the cleavage of the actin molecule , rather than interfering with the endogenous actin regulation pathways or by posttranslational modification of the actin molecule , to benefit its intracellular life cycle . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"intracellular",
"pathogens",
"viral",
"transmission",
"and",
"infection",
"molecular",
"probe",
"techniques",
"pathogens",
"cell",
"processes",
"microbiology",
"immunoblotting",
"fungi... | 2017 | A Legionella Effector Disrupts Host Cytoskeletal Structure by Cleaving Actin |
Biomphalaria glabrata snails that display either resistant or susceptible phenotypes to the parasitic trematode , Schistosoma mansoni provide an invaluable resource towards elucidating the molecular basis of the snail-host/schistosome relationship . Previously , we showed that induction of stress genes either after heat-shock or parasite infection was a major feature distinguishing juvenile susceptible snails from their resistant counterparts . In order to examine this apparent association between heat stress and snail susceptibility , we investigated the effect of temperature modulation in the resistant snail stock , BS-90 . Here , we show that , incubated for up to 4 hrs at 32°C prior to infection , these resistant snails became susceptible to infection , i . e . shedding cercariae at 5 weeks post exposure ( PE ) while unstressed resistant snails , as expected , remained resistant . This suggests that susceptibility to infection by this resistant snail phenotype is temperature-sensitive ( ts ) . Additionally , resistant snails treated with the Hsp 90 specific inhibitor , geldanamycin ( GA ) after heat stress , were no longer susceptible to infection , retaining their resistant phenotype . Consistently , susceptible snail phenotypes treated with 100 mM GA before parasite exposure also remained uninfected . These results provide direct evidence for the induction of stress genes ( heat shock proteins; Hsp 70 , Hsp 90 and the reverse transcriptase [RT] domain of the nimbus non-LTR retrotransposon ) in B . glabrata susceptibility to S . mansoni infection and characterize the resistant BS-90 snails as a temperature-sensitive phenotype . This study of reversing snail susceptibility phenotypes to S . mansoni provides an opportunity to directly track molecular pathway ( s ) that underlie the B . glabrata snail's ability to either sustain or destroy the S . mansoni parasite .
Schistosomes are parasitic trematodes that cause the chronic debilitating disease schistosomiasis , a neglected tropical disease that persists in over 70 countries of the developing world . It is estimated that at least 200 million people are chronically infected with the parasite with another 800 million remaining at risk for exposure . The disease burden is estimated at over 70 million disability-adjusted life years ( DALYs ) and there is increasing awareness that schistosomiasis can impact the epidemiology of other infectious diseases such as HIV ( especially in female patients with genital schistosomiasis ) . A concerted effort is , therefore , being made to develop novel intervention tools that include blocking transmission of the parasite at the snail stage of its life cycle [1]–[3] . Freshwater snails serve as obligatory intermediate hosts for the development of parasitic trematodes . Throughout South America and the Caribbean Islands the snail , Biomphalaria glabrata plays an important role in the transmission of Schistosoma mansoni . The relative ease of maintaining B . glabrata in the laboratory has enabled it to become the host/pathogen model system of choice in which studies aimed at elucidating the molecular basis of snail/schistosome interactions are being conducted . Thus far , studies using representative snail stocks that are either resistant or susceptible to the parasite provide an invaluable resource towards unraveling the complex biology of the snail/schistosome encounter . For example , using pedigree snail stocks with varying susceptibility phenotypes , a strong genetic basis was shown to exist for the susceptibility of B . glabrata to S . mansoni [4] . In adult B . glabrata , resistance to S . mansoni has been shown to be a dominant single-gene trait that is inherited by simple Mendelian genetics . In juvenile snails , however , genetics of resistance has been shown to be a complex trait , involving 5 to 6 genes each with multiple alleles . Similarly , genetics of susceptibility to the parasite either in juvenile or adult snails has been shown to be multi-genic [5] . Using snail stocks that represent these different susceptibility phenotypes , the genetic locus/loci governing these traits have been assessed by a variety of DNA genotyping tools . These studies have led to the identification of heritable markers that underscore the adult snail parasite resistant phenotype [6] . Advances have also been made towards the identification of genes associated with snail susceptibility phenotypes by examining differences in gene expression profiles between snails that are either resistant or susceptible in response to parasite infection [6]–[10] . Accordingly , several genes involved in the snail's innate defense system are now known to play a significant role in the balance of whether the snail becomes infected or not [11] , [12] . For example , in a resistant snail , such as the well-known representative BS-90 stock , the anti-parasite response in this snail has been shown to culminate in the encapsulation of the invading miracidia by a cell-mediated response involving hemocytes that , with plasma ( hemolymph ) factors , destroys the miracidium within a few days after it penetrates the snail . In a typical susceptible snail , such as the NMRI stock , however , there is no such active innate defense response against the invading miracidium and , therefore , the parasite survives , differentiates into sporocyts , producing cercariae that when released into freshwater can infect a human host , and go on to complete the life cycle . Aside from the well-recognized genetic basis of the snail-schistosome relationship , shared molecular determinants of both organisms ( snail and parasite ) are also thought to play a role in the snail host compatibility to S . mansoni . Thus , interactions of snail diversified fibrinogen-related proteins ( FREPs ) and polymorphic mucins of schistosomes have been identified as some of the target molecules of snail and parasite , respectively that either by interacting , or not , with each other define compatibility/incompatibility of the snail/schistosome encounter [13]–[15] . This concept of shared , or molecular mimicry , at the snail - parasite interphase , underlying mechanisms of schistosome-snail compatibility/incompatibility is referred to as the matched- mismatched hypothesis [16] . Variations in susceptibility of B . glabrata to S . mansoni have been well documented [17] , [18] . Furthermore , age- related variations in susceptibility have also been described . For example , Minchella and Richards showed that a snail that is susceptible as a juvenile can become resistant once it reaches adulthood , to the same strain of S . mansoni [19] . Given these variations , compounded with the fact that younger snails are , in general , more vulnerable to infection than adults [20] , we felt that to identify the mechanism ( s ) governing susceptibility to S . mansoni , in juvenile , rather than adult snails , might be more beneficial in the long run towards our eventual goal of blocking disease transmission in the snail host . For this reason , therefore , the present study was performed entirely with juvenile and not adult snails . To date , very few studies have investigated the modulation of stress genes and B . glabrata susceptibility to S . mansoni . However , Lockyer et al . ( 2004 ) , while examining differential gene expression between resistant and susceptible adult snails , in response to S . mansoni , detected upregulation of the transcript encoding the stress response gene , heat shock factor ( Hsp ) 70 in resistant but not susceptible snails after S . mansoni infection [21] . These results are in contrast to those we obtained [22] , showing instead upregulation of this transcript in early parasite exposed juvenile susceptible , but not resistant snails . Additionally , unlike the Lockyer et al . study where constitutive expression of Hsp 70 transcript was not observed in either resistant or susceptible adult snails , we showed that the expression of Hsp 70 occurs at similar levels in both normal resistant and susceptible juvenile snails . Furthermore , another study done using hemocytes collected from parasite exposed adult resistant and susceptible snails , showed down regulation of the Hsp 70 protein occurs in hemocytes of both these snails following infection , with more suppression of the transcript in susceptible than in resistant adult snails [23] . Thus , from the above studies , it is clear that there are major discrepancies concerning the expression of Hsp 70 between juvenile and adult resistant and susceptible snails , either with , or without infection , and also from hemocytes removed from infected adult resistant and susceptible snails . These discrepancies notwithstanding , as early as 1954 it was shown that raising the water temperature for maintaining B . glabrata shortened the length of the pre-patent period in S . mansoni infected snails , and also helped to maintain snail infectivity [24] . Additionally , this early study showed that some snails lost their infections when they were maintained at low temperature . In 1991 , Lefcort and Bayne showed that S . mansoni infected resistant snails ( 13–16-R1 stock ) displayed a preference for lower temperature compared to similarly exposed susceptible snails . No molecular explanations were , however , provided for these earlier observations [25] . While examining changes in gene expression profiles between juvenile resistant and susceptible snails soon after parasite exposure , we showed that the stress gene , Hsp 70 was induced early in susceptible but not resistant juvenile snails [9] , [22] . Subsequently , we showed that the Hsp 70 transcript was co-expressed with the transcript corresponding to the reverse transcriptase ( RT ) domain of the B . glabrata non LTR-retrotransposon , nimbus , after exposure of susceptible juvenile snails to normal but not to irradiated miracidia . Similar gene profiling studies done in B . glabrata after exposure to another trematode , Echinostoma paraensi , also reported the upregulation of Hsp 70 in response to this parasite infection in the snail [26] . Because of this apparent association of an early stress induction and juvenile snail susceptibility , in this study we tested the hypothesis that enhancing stress prior to infection of a representative resistant snail , such as the BS-90 stock , by non-lethal temperature modulation , reverses the resistance phenotype . The BS-90 snail originally isolated in the 1960s by Paraense and Correa in Salvador ( Brazil ) is a wild type snail that is resistant at any age ( either as juveniles or adults ) to both new and old world S . mansoni [17] . For this reason , most investigators have , since 1990 , used this stock for studies aimed at identifying genes that underlie the aforementioned active innate defense response seen in these snails against S . mansoni . At ambient temperature ( 25°C ) susceptible snails , such as the NMRI stock , reliably shed cercariae ( varying between 85–95% ) within 4 to 6 weeks after miracidia exposure , whereas exposed BS-90 snails destroy the parasite soon after infection , and thereby remain negative . Since arriving in our laboratory in 1990 , BS-90 snails have never been known to become susceptible . Furthermore , in an early series of experiments conducted by Paraense and Correa ( 1963 ) where these snails were exposed to S . mansoni under different temperature conditions during the coldest ( 19 . 5–22 . 7°C ) and warmest ( 24 . 9–27 . 6°C ) months in the laboratory , no effect of temperature was detected . Indeed , in this same study , the snails derived from the original stock ( collected in a lake near a beach at Amaralina district in Salvador ) , exposed to up to 100 miracidia remained negative . Here , we show that by subjecting the resistant BS-90 stock to non-lethal heat shock treatment at 32°C , herein referred to as heat-pulse , prior to exposure , the snails consistently reversed their phenotype , shedding cercariae 5 weeks after infection . In contrast , similarly exposed , but unstressed BS-90 snails , remained uninfected . Additionally , if the stressed BS-90 snails were immediately treated with the Hsp 90 inhibitor , geldanamycin ( GA ) before exposure , they remained resistant . Interestingly , treatment of the highly susceptible NMRI snails with 100 mM of the same inhibitor before exposure to S . mansoni also prevented infection . These findings are consistent with an apparent association of the induction of stress genes ( Hsp 70 , Hsp 90 and RT ) and B . glabrata susceptibility to S . mansoni . Furthermore , the temperature sensitive switching of the resistant phenotype of B . glabrata to S . mansoni susceptibility provides an important means of directly tracking mechanisms that underscore the parasite's survival or destruction in the B . glabrata intermediate snail host .
Female SW mice were purchased from Taconic ( Germantown , NY ) and maintained in the Biomedical Research Institute's ( BRI ) animal facility , which is accredited by Lewis et al . [27] , [28] the American Association for Accreditation of Laboratory Animal Care ( AAALAC; #000779 ) , is a USDA registered animal facility ( 51-R-0050 ) , and has an Animal Welfare Assurance on file with the National Institutes of Health , Office of Laboratory Animal Welfare ( OLAW ) , A3080-01 . Maintenance of the mice , exposure to S . mansoni cercariae , and subsequent harvesting of the adult worms were approved by the BRI Institutional Animal Care and Use Committee ( IACUC protocol approval number 09-03 ) . All procedures employed were consistent with the Guide for the Care and Use of Laboratory Animals . Juvenile snails were subjected to heat-pulse by incubation in pre-warmed ( 32°C ) sterile water for 1–4 hrs as previously described [22] . After the heat-pulse treatment , the snails were immediately exposed to S . mansoni miracidia ( 10 miracidia/snail ) at ambient temperature ( 25°C ) in fresh aerated tap water for at least 3 hrs , individually , as previously described [9] . The heat-pulse treatment of juvenile resistant snails was performed on the basis of our previous data that showed optimal induction of Hsp 70 and RT transcripts occurred in these snails only after prolonged ( 2 to 4 hrs ) heat stress [22] . Infected snails were maintained as described above , but in the dark . After 4 weeks , snails were screened individually for cercarial shedding by placing each snail in the well of a 12 well -plate ( ∼3 ml aerated water in each well ) under a light source at room temperature for 1 hr as previously described [22] , [31] . Given that the objective of this study was to examine the effect of temperature modulation on juvenile resistant snails , we scored snails as susceptible if they released any cercariae at all after infection , and not by how many parasites were shed per snail . This method of scoring was chosen to reflect the real life situation that it takes only a few viable parasites to transmit schistosomiasis in endemic regions . The water in individual wells was subsequently examined for the presence , or absence of cercariae under a dissecting microscope . The snails that were not shedding cercariae at the 9th week post-exposure , however , were kept and examined every week until the 12th week thereafter the snails were monitored on a monthly basis for cercarial shedding . After heat-pulse treatment as described above , BS-90 snails were transferred immediately into a beaker containing a solution of 100 mM geldanamycin ( GA ) ( Sigma Aldrich ) overnight at room temperature . Snails were removed from the drug solution , washed twice ( 5 min intervals ) in ∼30 ml of fresh aerated water at room temperature to rinse off residual contaminating drug before exposing to miracidia individually in 2 ml of fresh aerated water ( in a 12 well plate ) . Each snail was exposed in fresh water at room temperature while monitoring under a dissecting microscope for miracidia penetration . All snails were exposed to twice the number of miracidia ( 10 miracidia/snail ) we normally use for exposing juvenile NMRI snails . Different types of experimental conditions ( 12 snails for each cohort ) were set-up as follows: 1 ) normal ( unstressed and unexposed ) BS-90 snails , 2 ) parasite exposed-normal BS-90 snails , 3 ) heat-pulsed only ( unexposed ) BS-90 snails , 4 ) exposed then heat-pulsed BS-90 snails , 5 ) heat-pulsed/GA treated-exposed BS-90 snails , 6 ) GA treated normal BS-90 snails . Water changes were done on a weekly basis on all the snails . Snails were examined for evidence of parasite infection ( cercarial shedding ) as described above . In this study , aerated fresh water collected from the same container ( 30 gallons capacity barrel ) was used for all snail husbandry and miracidia exposures . To examine the effect of GA treatment on B . glabrata NMRI susceptible snails and parasite exposure , juvenile snails ( 4–6 mm in diameter ) were used for the study . Twelve snails in a cohort were treated with GA at final concentrations , 0 , 0 . 1 , 1 . 0 , 10 and 100 mM . Snails were treated as described above by incubating overnight in a 50 ml beaker containing the drug solution . Treated snails were removed from the drug solution , washed as described above before being transferred into 2 ml of aerated fresh water ( in a 12 well plate ) for individual miracidia exposure . Each snail was exposed in freshwater at room temperature ( 10 miracidia/snail ) and monitored under a dissecting microscope for miracidia penetration . After exposure , snails were transferred into fresh water for the remainder of the study . Experiments were repeated , 5 times , for the highest dose ( 100 mM ) of GA , and 3 times for the lower drug concentrations , representing 5 and 3 biological replicates , respectively . Twelve size-matched snails were used for each drug concentration . As a control , miracidia were treated with the same high dose of GA for 5 hrs then used for snail exposures as described above . Data were pooled ( 12 snails/dose of drug ) from 5 independent experiments for the high ( 100 mM ) concentration ( N = 60 ) and 3 independent experiments from the low ( 0 . 1–10 mM ) concentration ( N = 36 ) and standard error ( SE ) determined . In addition , three other controls ( 12 snails for each cohort ) were used as follows: 1 ) normal NMRI snails 2 ) exposed normal NMRI snails and 3 ) unexposed GA treated-NMRI snails . Water changes were done on a weekly basis on all the snails , and exposed snails were examined for evidence of parasite infection ( cercarial shedding ) as described above . From the 3 and 5 biological replicates ( N = 36 , and N = 60 , respectively ) , one-way ANOVA was used to determine if differences in cercarial shedding ( % ) between low and high doses of GA treatment of NMRI snails were significant . To investigate the expression of stress genes Hsp 70 , Hsp 90 and RT , snails from the same cohort , subjected to heat-pulse , and exposed individually to miracidia ( 10 miracidia/snail ) ( as described above ) were snap frozen in liquid nitrogen and kept at −70°C until required for RNA isolation . Of the three cellular stress genes examined in the present study , the differential induction of Hsp 70 and RT between susceptible NMRI and resistant BS-90 snails after parasite exposure have previously been reported [22] . To determine differences in the expression of Hsp 90 between juvenile BS-90 and NMRI snails they were exposed , individually as described above for 0 , 15 , 30 , 45 , 60 and 120 min before being frozen for RNA isolation . Total RNA was extracted from the whole snail by single-step simultaneous RNA isolation using RNAzol RT ( Molecular Research Center , Inc . , OH ) [32] . RNA from individual snails was utilized for all qPCR analysis . The quantity and quality of RNA was determined by UV absorbance ( A260 ) and an A260/A280 ratio ( ∼1 . 9–2 . 0 ) obtained by using the NanoDrop 1000 ( Thermo Scientific ) . Eighty nanograms of total RNA was analyzed by real time qPCR using Brilliant II SYBR Green QRT-PCR Master mix according to the manufacturer's instructions ( Stratagene , CA ) . Real time qPCR reactions were done using the 7300 real - time PCR system from ABI ( Applied Biosystem ) . At the beginning of the assay , a validation protocol ( done according to the manufacturer's instructions ) was performed to test amplification efficiencies of stress genes ( Hsp 70 , Hsp 90 and RT ) and myoglobin ( constitutively expressed house keeping gene ) [9] , [22] , [33] , [34] , [35] . The validation experiment was done using four different RNA templates dilutions to confirm that amplification efficiencies were equal between our genes of interest and myoglobin genes ( data not shown ) . The 25 µl final reaction volume contained either 200 nM of gene specific primers for Hsp 70 , Hsp 90 and RT or 50 nM of primers for the house keeping myoglobin gene that amplifies a 349 bp fragment [33] . Nucleotide sequences of the gene specific primers for Hsp 70 and RT have previously been described [22] . The Hsp 90 specific primer pair ( forward primer; 5′-tgtgcgcagagtgttcatcatgg-3′ and reverse primer; 5′-ctcctgtgaggcttcaatgagtc-3′ ) was designed by Primer-Blast software ( http://www . ncbi . nlm . nih . gov/tools/primer-blast/index . cgi ? LINK_LOC=BlastHome ) from publically available Expressed Sequence Tags ( ESTs of B . glabrata 5′end clone; accession number , AA547771 that has homology to heat shock protein 90 ) . The Hsp90 gene specific primers amplify a 199 bp fragment from B . glabrata BS-90 and NMRI stocks . All primers were checked for non-cross amplification with S . mansoni cDNA template as well as with other schistosome parasite species ( data not shown ) . Reactions were performed in triplicate ( technical replicates ) for each individual snail RNA sample in a one-step format with the first strand cDNA synthesis and real time PCR amplification done as described previously [35] . Each reaction contained a no template negative control to rule out non-specific amplification from contamination in the primers and buffers . The gene transcript levels were normalized relative to myoglobin expression . In our study , myoglobin showed stable expression in BS-90 and NMRI snails under normal and stress conditions ( data not shown ) . After five biological replicates using 10 snails/time point ( 50 snails in total ) the fold change in transcription of corresponding genes of interest were calculated by using the formula below [36] . P-values were calculated by comparing the delta-Ct value as previously described for each group by Student's t-test to determine if the differential expression of the transcripts between experimental and control groups was significant ( 9 , 22 , 32 , 33 ) . Fold-changes in transcripts corresponding to stress genes of interest , normalized relative to myoglobin expression from individual snail RNA ( 10 snails/time point , assayed in triplicate ) , were determined from five independent experiments . These data were pooled ( N = 50 ) and standard error ( SE ) determined . Each reaction contained a no template negative control to rule out non-specific amplification from contamination in the primers and buffers . To monitor the differential transcription of Hsp 90 between NMRI and BS-90 snails , RNAs were isolated at different time points following exposure , as described above , and were normalized relative to myoglobin expression . To determine constitutive expression of Hsp 90 in both BS-90 and NMRI snails , we performed qualitative RT-PCR as previously described [35] using first strand cDNA synthesized from total RNA using a cDNA synthesis kit ( Promega ) with M-MLV reverse transcriptase isolated from two same-sized ( 4–6 mm in diameter ) individual snails of either the resistant ( BS-90 ) or the susceptible ( NMRI ) stock . Using the Hsp 90 gene specific primers described above , second strand PCR was performed with cDNA template that was prepared either in the presence or absence of M-MLV RT under the following conditions: denaturation step at 95°C for 30 sec , annealing step at 65°C for 30 sec , extension at 72°C for 1 min , repeated for 29 cycles . The PCR products were resolved by agarose gel electrophoresis ( 1 . 2% ) and amplicons were visualized by ethidium bromide staining under UV trans-illumination using the Quantity One software ( Gel Doc XR imaging system , BioRad ) . A 100 bp ladder from a commercial source ( Invitrogen Laboratories , CA ) was utilized as standard for the qualitative analysis . Parallel PCR-amplification using equal amounts of the same cDNA template was done by using gene specific primers corresponding to the house keeping myoglobin gene as previously described [33] . PCR amplifications were also done without cDNA template , as negative control , to monitor the reagents utilized in the assay for possible contamination . Snails were screened as described above for cercarial shedding and those found to be positive at 7 weeks PE were allowed to shed into an appropriate container for 30–60 min under a light source at room temperature . After shedding , the snails were removed from the container and the cercariae suspension was allowed to stand ( 10 min at room temperature ) to allow any non-specific sediment ( mucoid material and feces ) from the snail to settle at the bottom of the container . The sediment-free top part of the cercariae suspension was transferred to another container and was used for the mouse infection . Two female Swiss-Webster ( SW ) mice ( 2–3 months old weighing18 grams ) were used for the cercariae infection ( 50 cercariae/mouse ) by tail exposure , allowing the cercariae to penetrate individually for ∼30 min . Forty-nine days after infection , feces from the mice were examined for first appearance of parasite eggs . The positive mice were euthanized for recovery of adult worms by the perfusion technique using 0 . 1 M citrate-0 . 15 M sodium chloride solution [28] . Mice infected with cercariae that were shed from the ‘resistant’ heat- pulsed -exposed BS-90 snails were perfused at 6 weeks post infection as described by Lewis et al . [28] . Harvested adult worms were frozen individually at −70°C until required . Genomic DNA was isolated from frozen individual worms as described by Simpson et al . [37] , and screened by RAPD-PCR using random primers OPR-14 and OPA-O6 ( EurofinsMWG/Operon , AL ) as previously described [6] . Amplified fragments were separated by agarose gel electrophoresis , and the presence/absence of bands visualized by ethidium bromide staining under UV Trans-illumination ( ChemiDoc imaging system , BioRad CA ) . For comparison , an equal amount of DNA from adult S . mansoni harvested from the susceptible snail was analyzed in parallel , using the same random primers mentioned above .
To investigate if the expression of Hsp 90 in the resistant ( BS-90 ) and susceptible ( NMRI ) snail would be similar as was shown previously for the levels of Hsp 70 and the nimbus RT domain , qualitative RT-PCR was conducted with template cDNA prepared with and without M-MLV-RT from RNA isolated from uninfected individual snails . As shown in Figure 1A , qualitative RT-PCR revealed that basal constitutive expression of Hsp 90 in both snail stocks was remarkably similar . Thus , the PCR amplification of cDNA from two individual BS-90 snails ( Fig . 1A , lanes 1 and 2 ) , using the Hsp 90 gene specific primers described in Materials and Methods , produced a 199 bp expected-size band that was similar in size and intensity ( Fig . 1A , lanes 1 to 4 ) to the PCR amplification product produced from cDNA templates from two individual NMRI susceptible snails ( Fig . 1A , lanes 3 and 4 ) . To rule out any possibility that the amplicons detected in lanes 1 to 4 might have originated from genomic DNA contamination in RNA preparations utilized for first strand cDNA synthesis , control reactions performed without reverse transcriptase ( −M-MLVRT ) in the first stand reaction were also utilized for PCR amplifications ( Fig . 1A , lanes 5 to 8 ) . Thus , using the Hsp 90 gene-specific primers for amplification reactions with this control ( minus M-MLV RT ) template produced no bands in either the BS-90 ( Lanes 5 and 6 ) or NMRI ( Lanes 6 and 7 ) samples , indicating that there was no genomic DNA contamination in the RNA preparations utilized for this assay . In parallel RT-PCR was performed with primers corresponding to the housekeeping myoglobin gene using equal amounts of the cDNA templates utilized in lanes 1 to 4 . In this case , ( also in lanes 10 to 13 ) an expected 350 bp size PCR product was obtained , indicating similar constitutive expression of myoglobin in both these two snail stocks . As another negative control , PCR was also conducted without cDNA template . The absence of a product ( lanes 9 and 18 ) showed there was no contamination in the reagents utilized for the assay . Since previous studies showed dramatic differences in the temporal modulation and degree of induction of the Hsp 70 and RT transcripts following S . mansoni infection between resistant and susceptible snails , in the present study we chose to examine whether parasite exposure mediates the differential expression of another major cellular stress gene , Hsp 90 . The induction of the Hsp 90 transcript in susceptible NMRI and resistant BS-90 juvenile snails exposed for different time points ( 0 , 15 , 30 , 45 , 60 and 120 min ) to miracidia is shown in Figure 1B . RNA isolated from the snails was analyzed by real time qPCR as described in Materials and Methods using the uniform constitutive expression of the myoglobin transcript in both snail stocks as internal standard . From 5 independent ( biological replicates ) assays done by using 10 individual snails per time point , with each RNA sample run in triplicate , results showed that as early as 15 min post- exposure almost 10-fold induction of the Hsp 90 transcript was obtained in the susceptible NMRI snail . Furthermore , the transcript remained upregulated in the susceptible snail throughout the 120 min PE time period examined . In contrast , a smaller difference in the upregulation of this transcript ( 1 . 54 fold change ) occurred in the resistant snail at the early 15 min time point after infection . Although variations in the induction level ( of Hsp 90 transcript ) were observed in the infected susceptible snail between the early 15–120 min PE time period , none of the levels we detected in the infected resistant snail ( 1 . 45 to 1 . 54 fold change ) exceeded those detected in the infected susceptible snail . Previously we ascertained , as mentioned above , that induction of stress genes occurs in the resistant BS-90 snail after a long 2–4 hrs time period either in response to parasite exposure , or heat shock . Accordingly , here we kept the BS-90 snails at 32°C for 3–4 hrs before exposing them immediately to miracidia as described in Materials and Methods . Using real time qPCR , modulations in expression of the three cellular stress genes ( Hsp 70 , Hsp 90 and RT ) were assessed . The fold changes in expression of transcripts corresponding to Hsp 70 ( Fig . 2a ) , Hsp 90 ( Fig . 2b ) and RT ( Fig . 2c ) in resistant BS-90 juvenile snails either following heat-pulse treatment , or subsequently subjecting the heat- pulsed snails to S . mansoni exposure for 2 hrs were determined ( Fig . 2 ) . As shown in Figure 2a , in comparison to snails exposed without prior heat-pulse treatment , minor induction ( 1 . 2 fold ) of the Hsp 70 transcript was detected in these resistant snails after parasite exposure . In contrast , however , the heat-pulse for either 3 or 4 hrs promoted a more significant induction ( 6 . 1 and 7 . 9 fold , respectively ) of this transcript . Interestingly , exposing the snails to miracidia immediately after 3 and 4 hrs heat-pulse kept the induced Hsp 70 transcript at a level that was higher ( 4 and 6 . 2 fold induction , respectively ) than the 1 . 2 fold induction we observed in snails that were exposed without being heat pulsed . In Figure 2b , while a 2 . 5 fold induction of the Hsp 90 transcript was observed in the unstressed 2 hrs parasite-exposed BS-90 snail , we detected a 7 . 9 and 7 . 8 fold increase in this transcript in snails subjected to 3 and 4 hrs of heat-pulse , respectively . The fold change in the Hsp-90 transcript remained relatively higher in snails that were heat-pulsed and immediately exposed to miracidia ( 3 . 3 and 2 . 9 fold increase ) compared to the 2 . 5 fold increase observed in the normal ( minus heat-pulse ) parasite-exposed snail . In Figure 2c , in unstressed BS-90 snails exposed to miracidia , a 3 . 5 fold increase in the RT transcript was observed compared to a 27 . 3 and 26 . 3 fold induction of this transcript when snails were kept for 3 and 4 hrs at 32°C prior to infection . Thus , in this case , the dramatic increase in the RT transcript remained elevated in snails that were heat -pulsed for 3 to 4 hrs and then immediately exposed to miracidia . Thus , a strong 20 . 5 and almost 10 fold increase was detected in these snails ( heat-pulse plus exposure ) compared to snails that were exposed without heat-pulse treatment where only a 3 . 3 fold increase in the RT transcript was observed . The above results showed that compared to snails that were only exposed to miracidia , the induction of all the stress transcripts examined ( Hsp70 , Hsp90 and RT ) was more elevated in BS-90 snails that were responding to parasite exposure after heat - pulse treatment . Results in Figure 3a shows the effect of combining heat -pulse treatment with immediate exposure of the stressed resistant snails to miracidia . Parasite-exposed unstressed snails were monitored as control . As shown in Figure 3a , as expected , juvenile resistant snails that were exposed to miracidia without heat -pulse treatment released no cercariae for the entire 7 weeks duration of the experiment . In contrast , a significant number , more than half ( 60 . 7% and 53 . 9% ) , of resistant snails that were heat -pulsed either for 3 or 4 hrs prior to exposure were found to shed cercariae starting at 5 weeks PE . Interestingly , by 7 weeks PE , all such prior heat -pulsed , and parasite -exposed resistant snails were shedding cercariae . To determine if cercariae released from these positive resistant snails were biologically viable i . e . capable of infecting the experimental mouse host and developing successfully into adult worms in this host , cercariae released from ‘resistant’ snails were utilized for mouse infections as described in Materials and Methods . Since we were able to perfuse adult worms from the mice infected with ‘resistant’-snail derived parasites , we can conclude that larval parasites released from these ‘resistant’ snails were indeed infectious , developing normally within the expected 6 weeks time-frame into adult worms in the infected mouse . To determine whether these adult worms were in fact genotypically identical to parasites harvested from mice exposed to cercariae released from our representative susceptible snails ( NMRI stock ) , we analyzed genomic DNA from worms harvested from mice that were infected by using either resistant , or susceptible snail derived parasites , with the multi-locus RAPD-PCR genotyping tool . As shown in Figure 3b , the DNA profile , using random primer OPR-14 , from 7 individual worms ( lanes 1 to 7 ) was comparable to that of an adult worm recovered from a mouse ( by perfusion ) that was exposed to cercariae shed from the representative susceptible snail ( lanes 8 to 10 ) . Likewise , DNA profiling of the same samples as utilized in lanes 1 and 7 but amplified , this time , with random primer OPA-06 showed no polymorphisms existed between either resistant ( lanes 1 to 7 ) or susceptible ( lanes 8 to 10 ) snail derived parasites . The absence of bands in lane 11 , using either of the two random primers but without template DNA shows there was no contamination in either the buffer or enzyme employed for the RAPD assay . As shown above , heat -pulse treatment followed immediately by miracidia exposure of the resistant snail , resulted in the robust resistant BS-90 snail becoming susceptible . On the basis of the above results , we , therefore , felt it was necessary to determine the outcome of exposing the resistant snails to miracidia before subjecting them to the heat -pulse treatment . Thus , juvenile BS-90 snails were initially exposed to miracidia , and immediately heat -pulsed at 32°C for 3 hrs . Figure 4a shows the real time qPCR results of the fold difference in expression of transcripts encoding Hsp 70 , Hsp 90 and RT at different time points ( 0 , 15 , 30 , 60 and 120 min ) after maintaining the infected snails at 32°C . Interestingly , unlike results shown in Figure 2 where enhanced induction of all three transcripts was detected in snails that were heat pulsed before exposure , in this experiment , induction of all the stress transcripts examined remained relatively unchanged . Accordingly , we monitored these snails ( parasite- exposed then heat -pulsed ) for cercarial shedding as described above . Results ( Fig . 4b ) showed that all ( 100% ) the resistant snails that were exposed to the parasite before being subjected to heat -pulse treatment remained negative . From these results , it is clear that exposing these resistant snails to the parasite before they are stressed , maintains the refractory status of these snails , further demonstrating the temperature sensitivity of BS-90 resistant phenotype to S . mansoni infection . To determine , more precisely , if significant upregulation of stress -related transcripts is indeed a factor in rendering the heat -pulsed resistant snails susceptible to infection , we investigated the effect of blocking the action of one of the stress proteins , Hsp 90 , by using a specific inhibitor drug geldanamycin ( GA ) that prevents this protein from performing its role as an essential chaperone in the cell [38]–[40] . In Figure 5 , resistant snails that were heat -pulsed for 3 hrs were exposed either immediately to miracidia as described above , or were treated with GA ( 100 mM ) before being exposed to miracidia . Results showed that without inhibitor treatment of stressed snails , the majority ( 70% ) of heat -pulsed and exposed snails were found to shed cercariae 9 weeks after exposure . In contrast , snails that were heat –pulsed , and immediately treated with GA before exposure failed to shed cercariae , remaining negative for the entire 9 weeks duration of the experiment . These data clearly show that blocking the action of Hsp 90 by GA treatment in the stressed resistant snails before exposure maintained their refractory phenotype , thereby indicating that a sustained significant stress induction of Hsp 90 was involved in the mechanism ( s ) of B . glabrata susceptibility to S . mansoni . To further demonstrate the link between stress induction in juvenile B . glabrata snails and their susceptibility to S . mansoni , here the effect of pre -treating the susceptible NMRI snail with the aforementioned Hsp 90 inhibitor ( GA ) on the outcome of infection was examined . Figure 6a shows the survival of snails after either infection alone ( minus GA inhibitor ) or after treatment with 100 mM of GA . Results ( Fig . 6a ) showed that snails tolerated the drug at this dose and survived at the same rate ( 100% ) as those that were exposed without drug treatment . Since all the snails tolerated this relatively high dose of the inhibitor , we proceeded to examine whether pre -treating the susceptible snail with various doses of GA would affect their susceptibility phenotype . As shown in Figure 6b , susceptible snails that were infected without prior drug treatment , as expected , were found to be shedding cercariae at 5 weeks PE ( 18 . 3% ) , with all untreated-exposed susceptible snails shedding cercariae at 9 weeks after infection . In contrast , susceptible snails that were treated with 100 mM GA prior to exposure failed to shed cercariae at the same 9 weeks PE time period while only a small percentage ( 3 . 7 to 4 . 8% ) of snails pre-treated with lower doses of GA ( 0 . 1 to 10 mM ) were found to be shedding cercariae . The longest surviving miracidial exposed , drug treated- ( 100 mM ) snails remained negative ( no cercarial shedding ) for up to 9 months after infection . All snails not shedding cercariae by the 9th week after exposure remained negative at week 12 and continued to be negative for at least 9 months PE . The reduction of shedding ( in percentages ) from infected snails treated with the higher dose of GA compared to the lower doses was significant as determined by one-way ANOVA ( P-value<0 . 05 ) [N = 60 and N = 36] . To rule out the possibility that results presented above might simply reflect the effect of the drug inhibiting the parasite's Hsp 90 homolog , thereby impairing the ability of miracidia to either penetrate the snail , or transform successfully into sporocysts , we treated miracidia directly with the highest dose of GA ( 100 mM ) that was utilized in this study . These drug -treated miracidia were then utilized for snail exposures in comparison to exposures done with untreated miracidia , as control . In these experiments , more than 75% of miracidiae ( with or without GA treatment ) penetrated the snails within 5 min , and all ( 100% ) successfully penetrated the snail within 1 hr ( data not shown ) . These data were similar to penetration behavior we previously observed where we used either normal or irradiated miracidiae for snail exposures [22] . As shown in Figure 7 , 41 . 7% and 87 . 5% of NMRI susceptible snails exposed to GA treated miracidia shed cercariae at 4–5 weeks . At 4 to 5 weeks PE , results analyzed by student's t-test showed that the percentage of cercarial shedding between NMRI snails exposed to either GA treated or normal miracidia was statistically significant ( P –value<0 . 05 ) . However , after week 5 PE , results showed that there was no statistically significant difference between the percentage of cercarial shedding between NMRI exposed to either GA-treated or normal miracidia . NMRI snails that were exposed to the untreated miracidia , likewise released cercariae after week 4 PE as we have routinely come to expect for snail infections performed by using this parasite strain and snail-host combination .
To date , very little information exists on molecular mechanisms that determine the outcome of the snail/schistosome interaction . Here , we have shown that upregulation of stress-related transcripts , such as those examined in this study , Hsp 70 , Hsp 90 and RT in the B . glabrata snail host , soon after infection , plays an important role in their susceptibility to S . mansoni . These results are consistent with our previous data that showed a differential induction of stress genes occurs between juvenile susceptible and resistant B . glabrata snails after exposure to S . mansoni miracidia . Thus , upregulation of transcripts corresponding to Hsp 70 and RT was detected sooner , and more dramatically in susceptible compared to resistant snails following either heat shock or parasite infection [22] . Also , in this previous study , we showed that the stimulus/stimuli for this stress induction may be present in normal but not in irradiated attenuated miracidia . Although we are yet to discover the nature of the parasite stress elicitor ( s ) , it is most likely released from the incoming parasite as excretory secretory products ( ESP ) . Interestingly , several studies have described using miracidial ESP to induce changes in either snail hemocytes , or the B . glabrata embryonic cell line , Bge [41]–[43] . Despite our limitation of not knowing what triggers the induction of stress in the snail ( soon after exposure to miracidia ) it is clear from this study that by using the heat -pulse regimen described to enhance the induction of stress to levels not typically seen under normal circumstances in resistant snails after exposure , it is possible to successfully reverse the resistant phenotype of juvenile resistant BS-90 snails i . e . render them susceptible to S . mansoni . Therefore , even though our results showed that parasite infection of the heat –pulsed snails caused a reduction in the induction of the stress transcripts , a more enhanced induction of all the three transcripts was still observed in heat-pulsed- infected snails than in snails that were exposed to miracidia alone without heat -pulse , helping to switch the phenotype of juvenile BS-90 snails from being resistant to susceptible . Thus , from these results , it is clear that resistance to S . mansoni in the juvenile BS-90 snail is a temperature-sensitive ( ts ) phenotype . Furthermore , cercariae released from these ts snails were infectious , developing fully into adult worms in the infected mouse , and with no change detected in DNA profiles of adult worms harvested from either these ‘resistant’ snails or our representative susceptible NMRI snail stock . Historically , studies regarding B . glabrata susceptibility to S . mansoni have emphasized either the role of genetics , or innate defense in the snail/parasite association . While these areas of study have been pivotal in explaining some of the complex dynamics behind why schistosomes are either destroyed or survive in the snail host , it is clear from our results that temperature sensitivity of the stress gene loci reported is a contributing factor in the outcome of this host/pathogen interaction . Accordingly , we can only speculate that the ts resistant BS-90 snail's phenotype must involve changes in the activation of the three stress response genes to account for their altered kinetics . In a recent study , we showed that an unknown external stimulus/stimuli from the parasite was indeed able to mediate the non-random repositioning of gene loci of interphase chromosomes in the snail embryonic cell line , Bge [44] . These gene loci repositioning studies have since been reproduced in intact snails responding to S . mansoni ( Arican , unpublished ) , and we are currently investigating if the stress genes are repositioned upon induction . Previous studies have shown that the induction of stress is important for successful outcomes of other host-pathogen relationships as well . For example , in baculovirus infected Sf-9 cells , an increase in the expression of Hsp 70 was found to correlate with active virus replication . Thus , in this study , Lyupina et al . showed that inhibiting Hsp70 expression by the drug KNK437 suppressed virus replication [45] . Additionally , Hsp 90 has been shown to be essential for the growth of the malaria-causing agent , P . falciparum , in human erythrocytes [46] . Consequently , derivatives of GA , the Hsp 90 inhibitor used in the present study , has also been used to inhibit the growth of P . falciparum and another protozoan , Trypanosoma evansi [46] , [47] . GA is a benzoquinone ansamycin antibiotic that binds to the N-terminal ATPase site of Hsp 90 to inhibit its chaperone activity . Although the inhibitor and its derivatives have previously been used to inhibit cell proliferation in cancer [48] and the growth of other parasites , as mentioned above it has never been shown , until now to treat any mollusk in the context of examining host-pathogen interactions . Interestingly , our results showed that GA treatment neither impaired the penetration behavior of the miracidia nor their ability to remain infectious . More studies , especially performed in vivo for longer time points will be needed to further examine this apparent lack of GA toxicity on the larval parasites . A similar lack of GA toxicity , which might be due to non-binding of GA to Hsp 90 homologues of free-living nematode larval stages , has previously been reported [49] . Heat shock proteins are highly conserved proteins that have been shown to play a critical role in maintaining protein integrity , preventing the aggregation of misfolded proteins in the cell , thereby maintaining normal cell function in the face of cellular injury from physical or physiological stress [50] . Very few studies have examined stress induction in relation to mollusk/pathogen interactions . Our results showing the very early ( within 15 min ) induction of Hsp 90 in susceptible snails ( but not resistant snails ) after infection was surprising and underscores the need for more studies on this stress protein , especially in relation to snail-schistosome interactions . Hsp 90 is expressed abundantly even in the absence of stress and constitutes a large portion of constitutively expressed protein in cells . The protein is regarded as being essential to cell viability because of its central role in forming complexes with a wide variety of co-chaperones and client proteins that are involved in major cellular pathways , such as signal transduction and cell-cycle control [51] . How Hsp 90 interacts directly , or indirectly with either the B . glabrata Hsp 70 or nimbus RT has yet to be investigated . Particularly , whether ( or not ) key molecules of the snail's innate defense system , such as FREPs are client proteins of Hsp 90 , remain to be investigated . Previously , we showed that co-induction of Hsp 70 and nimbus RT transcripts occur soon after S . mansoni infection of juvenile susceptible snails [22] . Mobile Genetic Elements ( MGEs ) , such as nimbus are responsive to cellular stress [52] , [53] . However , the role of the nimbus non-LTR retrotransposon in the stress pathway of B . glabrata remains unknown . Further studies are , therefore , required to elucidate the relationship between all these stress genes ( Hsp 90 , Hsp 70 and RT ) in the snail's behavior towards S . mansoni . In other mollusks , such as the clam , Mercenaria mercenaria , the upregulation of Hsp 70 was observed in this mollusk in response to the opportunistic parasite , known as Quahog Parasite Unknown , QPX [54] . Also , in another clam , Meretrix meretrix , it has recently been shown that expression of Hsp 70 was upregulated soon after Vibrio parahaemolyticus infection [55] . Additionally , in the disk abalone , Haliotis discus , a recent molecular characterization showed that Hsp 90 is induced within 4 hrs after treatment with lipopolysaccharide , LPS [56] . In another B . glabrata susceptible snail , the M-line stock , Hanington et al . [26] showed upregulation of stress related transcripts following infection of these snails with trematodes , either S . mansoni or Echinostoma paraensei . The modulation ( down regulation ) of Hsp 70 in hemocytes isolated from exposed B . glabrata resistant and susceptible snails ( the cells most intimately associated with the active destruction of schistosomes in the snail host ) has also been reported , suggesting an involvement of this stress protein in the snail host's defense system [7] , [23] . Indeed , an immunological role for stress proteins has been widely documented [57] . Thus , it might be reasonable to assume that in the snail/schistosome system , cellular stress triggered against parasite proteins that are recognized in the snail as non-self , by maintaining the homeostasis of the host , paradoxically protects the parasite as well . Larval schistosomes have been shown to express RNA transcripts for heat shock proteins . Presumably , such heat shock proteins ( released from the parasite ) might induce stress genes in the snail host , providing the cytoprotection that the parasite needs for its own successful invasion . While a strong anti-schistosome Hsp 70 humoral response has been reported in several infected ( S . mansoni and S . hematobium ) mammalian ( murine , human and baboon ) hosts [58] , [59] nothing is known about the role of schistosome heat shock proteins and the snail's innate defense system . Thus far , we have evidence showing that an active defense system plays an important role in the BS-90 resistant snail's ability to ward off the parasite infection . By showing in this study that the deliberate use of stress in the form of non-lethal heat-pulse ( boosting the level of inducible stress in the resistant BS-90 snail before infection ) was a necessary step in rendering these normally robust resistant snails susceptible , we can suggest that the stress induced dampened the anti-schistosome response that is typically seen in these snails . Previously , it was shown that the resistance phenotype can be interfered with in resistant snails ( 10-R2 and 13–16-R1 stocks ) if snails were first infected with other trematodes , such as E . paraensei and E . lindoense before being exposed to S . mansoni [60] . While no molecular explanation was given for this apparent suppression of the defense system by the dual infection protocol first described by Lie and Heyneman et al . [61] , these early results showed that susceptibility to S . mansoni in these resistant stocks developed shortly ( within 1 hr ) after they had been exposed to E . paraensei . In light of our current results , we can assume that in this previous study , the primary echinostome infection , by triggering a stress response , dampening the innate defense system allowed the secondary S . mansoni infection to survive and develop . In other host-pathogen systems , there is clear evidence that expression of stress proteins , in particular Hsp 70 is an important feature in modulating the host innate immune response [62] . Another plausible explanation for the results presented here might be that the initial heat -pulse could have destroyed the resistant snail's hemocytes , thereby rendering these cells incapable of killing the incoming miracidia . It is also possible that the induction of stress might be a reflection of the successful establishment of the parasite in the snail . As mentioned above several factors , and not hemocytes alone govern compatibility/incompatibility issues between B . glabrata and S . mansoni . The heat pulse regimen utilized to enhance the induction of stress genes in the resistant snails was not lethal . All the snails survived at this elevated temperature , and a colony of BS-90 snails that we maintain ( now in their eighth month ) at 32°C continue to thrive . Interestingly , all ( 100% ) of progeny snails ( F1 , exposed to miracidia and kept at 32°C after exposure ) bred from BS-90 snails maintained at 32°C were found to shed cercariae at 3 to 4 weeks PE ( Ittiprasert , Miller and Knight unpublished ) . While we have no data supporting the notion that higher prevailing environmental temperatures might facilitate snail susceptibility , our results show that it is possible that climate change might impact resistant snail susceptibility to schistosomes . Indeed , a recent study showed that global warming might result in an increase in cercarial output of infected snails [63] . By using a recently developed gene silencing method based on soaking snails in siRNA coupled to the inert cationic carrier , polyethylene imines ( PEI ) [44] , we are currently working to systematically knock-down transcription of Hsp 70 , Hsp 90 and nimbus RT in the snail by the PEI delivery tool . Thus far , preliminary results indicate that knocking down these transcripts to levels comparable to those we routinely obtain for suppressing the expression of low copy RNA transcripts will be more challenging . Despite these initial challenges , we hope to elucidate the role of these stress proteins in the snail host schistosome relationship by knocking down their corresponding transcripts . In addition , we have used the Hsp 90 inhibitor drug used in this study , GA , for treating pre-patent ( 2 week exposed snails ) and results show consistently that once established , the drug has no effect on the infection and all treated pre-patent snails go on to shed cercariae . In conclusion , we have shown in this study that by applying stress in the form of mild heat pulse to resistant BS-90 snails before they are exposed to S . mansoni , renders these snails susceptible . In contrast , infecting these snails before stressing them does not reverse their resistance phenotype , suggesting that the stress induction is an early necessary step in the sequence of molecular events that contribute towards making a snail susceptible . In addition , use of the stress inhibitor to treat susceptible snails before exposure was able to prevent them from shedding cercariae , again confirming that the stress pathway is indeed required for snail's to succumb to the parasite infection . These data open up a new opportunity to delve into unraveling the mechanism ( s ) that helps snails to either overcome or sustain the S . mansoni parasite infection , investigations that should lead to developing novel tools to interfere with schistosome-snail infections and thus reduce transmission of schistosomiasis . | Biomphalaria glabrata snails that are either resistant or susceptible to the parasite , Schistosoma mansoni , have been an invaluable resource in studies aimed at understanding the molecular basis of the snail/schistosome interaction . Schistosomes cause the chronic debilitating disease schistosomiasis . Thus , it is hoped that dissecting pathways that underlie the snail/schistosome relationship might translate into alternative control strategies that will include blocking transmission of the parasite at the snail-stage of its development . Induction of stress genes is a feature distinguishing early exposed juvenile susceptible NMRI snails from resistant BS-90 snail stocks . To further analyze this apparent involvement of stress induction and snail susceptibility , here we applied heat stress to the resistant BS-90 snail , enhancing induction of stress genes ( Hsp 70 , Hsp 90 and RT ) prior to infection . Results showed these resistant snails became susceptible , indicating resistance as being a temperature sensitive phenotype in these snails . Stressed resistant snails treated with the Hsp 90 specific inhibitor , geldanamycin , prior to exposure , were , however , shown to maintain their refractory phenotype . Interestingly , inhibitor treated susceptible snails also became non-susceptible . Collectively , these data point to stress induction as an important early step in the ability of S . mansoni to infect juvenile B . glabrata snails . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"systems",
"biology",
"genetics",
"immunology",
"biology",
"zoology",
"genetics",
"and",
"genomics"
] | 2012 | Reversing the Resistance Phenotype of the Biomphalaria glabrata Snail Host Schistosoma mansoni Infection by Temperature Modulation |
Early studies have shown that moderate levels of calcium overload can cause lower oxidative phosphorylation rates . However , the mechanistic interpretations of these findings were inadequate . And while the effect of excessive calcium overload on mitochondrial function is well appreciated , there has been little to no reports on the consequences of low to moderate calcium overload . To resolve this inadequacy , mitochondrial function from guinea pig hearts was quantified using several well-established methods including high-resolution respirometry and spectrofluorimetry and analyzed using mathematical modeling . We measured key mitochondrial variables such as respiration , mitochondrial membrane potential , buffer calcium , and substrate effects for a range of mitochondrial calcium loads from near zero to levels approaching mitochondrial permeability transition . In addition , we developed a computer model closely mimicking the experimental conditions and used this model to design experiments capable of eliminating many hypotheses generated from the data analysis . We subsequently performed those experiments and determined why mitochondrial ADP-stimulated respiration is significantly lowered during calcium overload . We found that when calcium phosphate levels , not matrix free calcium , reached sufficient levels , complex I activity is inhibited , and the rate of ATP synthesis is reduced . Our findings suggest that calcium phosphate granules form physical barriers that isolate complex I from NADH , disrupt complex I activity , or destabilize cristae and inhibit NADH-dependent respiration .
The heart is one of the most energy dependent tissues in the body and requires a very high rate of ATP synthesis and oxygen delivery . Energy stored in carbon fuels is extracted by a network of biochemical reactions to produce reducing equivalents such as NADH and UQH2 . These reducing equivalents feed electrons into the electron transport system to generate a proton electrochemical gradient which is then utilized to synthesize ATP . This entire process is tightly regulated and maintains stable levels of ATP despite rapid changes in workload and energy demand [1] . But when the heart is subjected to stressors , such as ischemia , this tight coupling quickly becomes destabilized . Rhythmic cytosolic calcium signals characterized by spike-like transients occur during normal physiological function in the heart . These calcium transients increase peak cytosolic calcium levels from approximately 100 nM up to levels ranging from 500 nM to 1 μM , depending on energy demand [2] . These transients can result in net calcium accumulation by mitochondria via the calcium uniporter complex [3 , 4] and cause matrix calcium levels to increase from 100 nM to 1 μM and stimulate NADH and ATP production rates [5–7] . To prevent the accumulation of toxic levels of calcium , the mitochondrial sodium calcium exchanger ( NCLX ) [3 , 8 , 9] removes one matrix calcium ion in exchange for three cytosolic sodium ions in an electrogenic exchange mechanism . Under ischemic conditions , cytosolic calcium levels can rise up to 3 μM [10] . At these high levels of cytosolic calcium , calcium uptake exceeds calcium efflux [11] . This can translate into massive increases in mitochondrial calcium content reaching up to 1 M total calcium stored as calcium phosphate granules [12–15] . In this calcium overloaded state , the ATP production capacity of mitochondria is severely compromised [16] , and mitochondria are primed for a devastating injury upon reperfusion known as mitochondrial permeability transition [17–19] . The exact mechanism behind the reduction in mitochondrial ATP production rates caused by calcium overload is unknown . Some have argued that calcium overload results in a direct inhibition of complex I [20 , 21] . This is supported by data that show complex I activity is reduced after reperfusion injury [22] . Others have reported that calcium overload inhibits the matrix dehydrogenase complexes pyruvate dehydrogenase and alpha-ketoglutarate dehydrogenase [23–25] . In addition to these reports , others have claimed calcium overload directly inhibits the adenine nucleotide translocase [26] , reduces the availability of free and/or Mg-bound ADP for oxidative phosphorylation and transport caused by calcium chelation [27 , 28] , causes net loss of matrix purine nucleotides [29] , or lowers the membrane potential for ATP production [30] . Another study points towards inhibition of cytochrome c oxidase by high levels of calcium [31] . Calcium phosphate granules have also been implicated in the inhibition of oxidative phosphorylation [32] . Yet , in other studies , mitochondrial calpains have been suspected to play a major role in calcium-induced mitochondrial dysfunction [33 , 34] . Needless to say , the cause of impaired mitochondrial function due to calcium overload still remains uncertain . In this study , we utilize a joint experimental and computational approach to identify the likely mechanisms and gain physiological insight into the effects of calcium overload on mitochondrial function and ATP production capacity . We tested the effect of low to moderate calcium overload conditions on ADP-stimulated respiration and membrane energization . The levels tested were below the threshold sufficient to cause mitochondrial permeability transition and demonstrate a permeabilization-independent effect of excess calcium on mitochondrial function . We found that calcium chelation or depressed membrane potential during sodium/calcium cycling did not lead to lower rates of oxidative phosphorylation . In addition , we did not detect any loss of purine nucleotides or direct inhibition of either the adenine nucleotide translocase or cytochrome c oxidase . In addition , our results show that mitochondrial calpains are not the cause of this phenomenon . Our data support the idea that complex I inhibition is the most probable cause of the observed decrease in rates of oxidative phosphorylation . And it is the amount of calcium phosphates accumulated by mitochondria , not the matrix free calcium concentration , that is the key determinant of this inhibition .
Based on the data shown in Fig 1 , the following mechanisms explain the calcium inhibition data: i ) lower membrane potential caused by sodium/calcium cycling lowers the driving force for ATP production; ii ) a subpopulation of mitochondria permeabilize and reduce the total number of mitochondria capable of synthesizing ATP; iii ) calpain activation leads to proteolytic loss of electron transport and/or ATP synthesis function; iv ) available ADP and/or phosphate for phosphorylation is depleted or reduced by calcium binding , incorporation into calcium phosphate granules , or net loss; and iv ) direct inhibition of calcium on mitochondrial processes critical for ATP production . We employed both experimental and computational strategies to test these hypotheses . The first hypotheses states that the lower membrane potential caused by sodium/calcium cycling lower the thermodynamic driving force , and hence , the rate for ATP synthesis . And while the membrane potential , as shown in Fig 1B , is lower after the CaCl2 bolus was given due to sodium/calcium cycling , the membrane potential during ATP synthesis is the same for all calcium conditions . This rules out that calcium-dependent reduction in respiration rates during ADP-stimulated respiration is due to a lowering driving force for ATP production . Thus , hypothesis i ) is disproved . The second hypothesis involves subpopulations of mitochondria responding differently to the calcium challenges . Isolated mitochondria in suspension do not undergo calcium-dependent permeabilization all at once . It is believed that the individual calcium tolerance of each mitochondrion in a population of suspended mitochondria falls within a distribution [35] , so that the more susceptible mitochondria undergo permeability transition first . Upon permeabilization , they release their calcium content which is taken up by more calcium-resistant mitochondria until they reach their calcium limit , and so on . This cascade of events might explain the data given in Fig 1A . However , both the membrane potential data in Fig 1B and the calcium uptake data in Fig 1C dispute this notion . The membrane potential data reveal that energized status of the population of mitochondria is stable , and the calcium uptake data show that calcium uptake is robust with no detectable permeabilization . This disproves hypothesis ii ) . The next hypothesis argues that calpain activation is the reason why calcium lowers ADP-stimulated respiration . Calpains are intracellular calcium-activated cysteine proteases present in nearly all vertebrate cells [36] . They are often , but not always , associated with subcellular organelles . The physiological function of calpains are not fully understood but they likely play major roles during autolysis . Calpains remain inactive until calcium increases to sufficient levels . In the ischemic myocyte , the dysregulation of cytosolic and mitochondrial calcium is thought to be central to calpain activation . Studies have shown that mitochondrial calpain activation does cleave specific sites on complex I and ATP synthase which has been linked to poor substrate oxidation and ATP production [33 , 34 , 37] . So we decided to test whether or not calpains played any sort of role in the calcium-induced depression of ADP-stimulated respiration . We experimentally tested this hypothesis by incubating isolated mitochondria with calpain inhibitors prior to calcium exposure . If calpains were responsible for this phenomenon , then inhibiting their proteolytic activity should rescue the observed calcium triggered mitochondrial dysfunction . Our results shown in Fig 2 demonstrate that calpain inhibitors do not relieve this inhibition and thus are not involved in the observed phenomenon . Longer incubation times did not improve mitochondrial function ( S1 File ) . Hypothesis iii ) is disproved . In order to test the remaining hypotheses , we resorted to utilizing a computational approach . We started with the bioenergetics model from Bazil et al . [38] and added calcium handling , matrix calcium buffering , and updated a few other reactions to improve fits to the original data [39] . See the Supporting Information ( S1 File ) for details . In brief , very simple mitochondrial calcium uniporter ( MCU ) and sodium calcium exchanger ( NCLX ) rate expressions were used with the sodium hydrogen exchanger rate expression from Bazil et al . [40] . We first started with more biophysically detailed MCU [41] and NCLX [42] rate expressions; however , we encountered several issues which precluded us from incorporating these more detailed expressions into the current model . First , the detailed MCU rate expression showed saturation with respect to respiration . This aberrant behavior was due to a calcium dissociation constant for the MCU that was too low causing the calcium current to saturate and produce a shark fin like respiratory dynamic . From Fig 1A , it can clearly be seen no such saturation-like respiratory response is evident . As such , we used a much simpler rate expression with a higher Vmax and KD for calcium that is more in line with the known characteristics of the channel [43] . In addition to the MCU , the detailed NCLX flux expression was overly complex and prevented adequate fits to the data . In this case , the steep proton dependency was determined to be the main problem . As with the MCU problem , a simpler rate expression was adequate to fit the experimental data . We also included the model of the mitochondrial calcium sequestration system based on data from Blomeyer et al . [44] described in detail below . The last major changes to the model was to the lumped TCA cycle rate expression and updating the complex III rate expression . We noticed the rate expression published with the original model failed to adequately simulate the respiratory behavior after calcium loading . This was due to calcium-dependent alkalization of the matrix not being properly compensated for in the expression and was alleviated by slightly altering the NADH feedback component . In addition to changing the TCA cycle rate expression , we replaced the complex III rate expression with a new one that includes dimer functionality from Bazil et al . [45] . The new rate expressions are given in the Supporting Information ( S1 File ) . Fig 3 shows the updated schematic of the bioenergetics model with new rate expressions and components , the new updated model fits to the experimental data it was originally parameterized on , and the parameter correlations shown as a heat map for the adjustable parameters given in Table 1 . As shown in the center four panels in Fig 3 , the model fits the data equally well as the original model , and in several cases even better . In particular , the updated model fits are more faithful to the data for the NADH and ADP data . The parameter correlation heatmap shown in the right panel of Fig 3 reveals the model parameters are relatively uncorrelated except for strong correlations between the MitoDH and electron transport related parameters . Local sensitivity coefficients are normalized and averaged with all model outputs aligned with the experimental data . Only non-zero sensitivity coefficients were averaged . The leak activity for the model simulations of the guinea pig data ( calcium-dependent inhibition data ) was lowered by 40% to account for the tighter respiratory coupling that these mitochondria possess . There is lot of evidence that calcium is reversibly sequestrated as calcium phosphate granules to maintain energy homeostasis [12 , 13] . Although we have implicitly modeled granule formation and calcium storage using an empirical model [46] , this approach required an artificially high proton buffering capacity to prevent excess matrix alkalization . This over alkalization was corrected by explicitly modeling calcium phosphate granule formation and including proton release as a part of this process [13] . A general reaction formula for calcium phosphate formation is depicted in Eq 1 . This equation is charge and atom balanced and works for many types of calcium phosphate stoichiometries including tricalcium phosphate , hydroxyapatite , and octacalcium phosphate . In the case that 3m-2n < 0 , the H in the calcium phosphate complex is understood to be OH with a stoichiometry of 2n-3m . Assuming an additional prototypical buffering component , the total matrix calcium buffering power that includes the formation of calcium phosphate is shown in Eq 2 . In Eq 2 , we use the concentration of hydrogen phosphate , as opposed to the phosphate ion , for maximum compatibility and simplicity . The bioenergetics model does not include the phosphate ion as a state variable because this ion only exists in appreciable quantities under very basic , non-physiological conditions . Including the phosphate ion would not change the results but require addition state variables and parameters to be added to the model . Fig 4 shows the relationship between mitochondrial calcium buffering power and matrix free calcium with the explicit representation of calcium phosphate granule formation compared to the experimental data from Blomeyer et al . [44] . The model consists of two essential components . The first component characterizes prototypical calcium buffering caused by binding to proteins and lipids not explicitly accounted for in the model . The second component represents the formation calcium phosphate granules . Table 2 gives the parameters used to simulate the data given in Fig 4 . The calcium phosphate complexation constant is not a solubility product constant . It is a parameter that governs the matrix calcium buffering power . For this data , either tricalcium phosphate , hydroxyapatite , or octacalcium phosphate formation fit the data with near equality . As shown in the inset of Fig 4 , the major difference between the three types of calcium phosphates is that higher buffering powers for matrix calcium concentrations exceeding 10 μM are obtained as calcium stoichiometric coefficient increases . For simplicity , we chose to model tricalcium phosphate formation . In addition , we assume calcium phosphate precipitate formation is rapid relative to calcium uptake and efflux . Further experiments are required to determine which calcium phosphate species constitute the main precipitate . We then proceeded to simultaneously fit the model to the data shown in Fig 1 by allowing some of the calcium related parameters to change . For these simulations , we ignored the calcium-dependent reduction in ADP-stimulated respiration to get a baseline model capable of simulating the calcium handling data . To minimize the number of adjustable parameters , we selected the most important parameters with respect to calcium handling using sensitivity analysis conducted with estimated parameter values obtained from the literature . These parameters are given in Table 3 and were determined to be the MCU activity , MCU calcium dissociation constant , NCLX activity , and the NCLX calcium dissociation constant . All other model parameters used for the simulations are given in the Supporting Information ( S1 File ) . The most sensitive parameter is the MCU activity . This parameter has a high , normalized local sensitivity coefficient relative to the others . The next highest ranked sensitivity is for the MCU calcium dissociation constant . The remaining two parameters , the NCLX activity and the NCLX calcium dissociation constant have near equal sensitivity values . The parameters were relatively correlated ( between 0 . 7 and 1 ) which indicates additional information is needed to uniquely identify these values . Also , the NCLX parameters are dependent on the matrix calcium buffering parameters given in Table 2 . However , these correlations do not impact the model results in this study . Local sensitivity coefficients are normalized and averaged with all model outputs are aligned with the experimental data . Only non-zero sensitivity coefficients were averaged . With the model calibrated , we began to computationally test the remaining hypotheses . Fig 5 shows that the reduction in ADP-stimulated respiration is not due to a reduction in available ADP and/or phosphate for oxidative phosphorylation . Even the highest calcium bolus only leads to less than 1% and 2% reduction in available ADP and phosphate , respectively . This makes it highly unlikely that this is the cause of the calcium-dependent reduction in respiration shown in Fig 1 . However , these simulations do not rule out the possibility that ADP is physically incorporated into the calcium phosphate granules . That remains a possibility to be tested , thus the hypothesis is not fully disproven . We will revisit this hypothesis below . We next tested the remaining hypothesis by adding a calcium-dependent Michaelis Menten-like inhibitory function to the main reactions involved in ADP-stimulated respiration . We opted to use a simple inhibitory function as opposed to more non-linear Hill type functions for simplicity . For this , we fit the adjustable parameters given in Table 3 to simulate the calcium handling components of the mitochondrial bioenergetics model for the experimental conditions described in Fig 1 . All other parameters were fixed and given in detail in the Supporting Information ( S1 File ) . Table 4 summarizes the results and shows that the best mechanism that matches the data is the inhibition of complex I as a function of the concentration calcium phosphate granules . The next best mechanism is the inhibition of complex III as a function of calcium phosphate granule concentration . All other conditions did not lead to adequately good fits to the data based on the normalized least squares values and visual inspection . The relatively small inhibition constants for the inorganic phosphate carrier and F1FO ATP synthase are expected , as these rate expressions have high activities to put them in a near equilibrium state . More complex inhibition functions involving multiple reactions were not considered . These results show that the most likely site of inhibition is at complex I or III and the calcium phosphate granule concentration is the likely cause . Fits were quantified using a normalized least squares value which was computed by summing the square of the difference between model and data relative to the data and divided by the total number of data points . The model results using the complex I inhibition mechanism compared to the experimental data is shown in Fig 6 . As shown , the model faithfully reproduces the experimental data . The model captures the calcium-dependent stimulation of respiration caused by sodium/calcium cycling and the inhibitory effect of calcium phosphate precipitates on ADP-stimulated respiration . In addition , the model reveals why the mitochondrial membrane potential is more polarized during leak state respiration when 1 mM EGTA is present . With the model , we can mechanistically explain this observation . The reason why the membrane potential is higher in this respiratory state when 1 mM EGTA is present , the matrix pH is closer to the buffer pH . This lowers the ΔpH component of the proton motive force and leads to an increase in the membrane potential . The matrix alkalization when 1 mM EGTA is not present is due to uptake of the residual calcium in the buffer . This residual calcium contaminate comes from reagent impurities mainly coming from potassium chloride and potassium phosphate salts . This is a small , but important , validation of the model ability to accurately simulate mitochondrial physiology . We used the model results given in Table 4 to devise an experiment to further rule out as many of the model-generated hypotheses as possible . Since all the reactions given in Table 4 have some dependence on NADH-dependent reactions , we chose to perform a similar experiment shown in Fig 1 but use succinate as the fuel to bypass any NADH-dependent reactions . The results from those experiments are shown in Fig 7A . To simulate the experiments , the two parameters from the empirical rate expression in Bazil et al . [38] for succinate-dependent respiration was modified to match the leak state and ADP-stimulated respiratory rates when 1 mM EGTA was present . For details , see the Supporting Information ( S1 File ) . While the calcium boluses result in a minor decrease in ADP-stimulated respiration , the magnitude is far less than that of ADP-stimulated respiration using NADH-linked respiratory fuels . As can be seen , the model faithfully reproduces the experimental data supporting the hypothesis that calcium phosphate granules lead to complex I inhibition which lowers the ADP-stimulated respiration rates in a titratable manner . These results dispute the notion that ADP is sequestered in these calcium phosphate granules because the ADP-stimulated respiration rates are hardly reduced . These results also rule out the complex III inhibition mechanism as the primary mechanism explaining the calcium-dependent inhibition of ADP-stimulated respiration , and only leaves calcium phosphate dependent inhibition of complex I .
Ischemic myocardium is characterized by a chronic elevation in cellular calcium concentration . In this setting , mitochondria accumulate calcium to levels that compromise their ability to synthesize ATP . The calcium accumulated during ischemia is mostly present in the form of insoluble calcium phosphate granules [12–15] . Our experimental results show that calcium levels below thresholds that trigger mitochondrial permeability transition significantly lowers the rate of ADP-stimulated respiration . Some studies from rat heart , liver and brain mitochondria suggest that matrix calcium overload can directly inhibit ANT activity and reduce the ATP synthase rate [28 , 29 , 47] . Other studies demonstrate that the flux of pyruvate dehydrogenase complex can also be inhibited with matrix calcium overload [23] . In addition to these studies , there are reports that calcium stimulates cytochrome c dissociation from the membrane and alters complex III and IV activities [48] . Still others link calcium overload and complex I activity so that in the face of elevated levels of calcium , the ATP synthase rate is significantly reduced [20 , 22] . Calcium-dependent matrix proteases have also been implicated in calcium-induced mitochondria dysfunction [33 , 34] . These calcium effects on mitochondrial metabolism are not necessary mutually exclusive . There could be a spectrum of calcium overload responses that depend on tissue , environment , and extent of ischemia/reperfusion injury . To understand the underlying mechanism of calcium inhibition on ADP-induced respiration , we combined both experimental and computational approaches using the classic model-based design of experiments paradigm . The experimental approach utilized the use of isolated mitochondria and interrogated the respiratory , energetic , and calcium handling response to various boluses of calcium and mimic the calcium overload conditions that occur during ischemia . These responses were quantified and used to inform a computational model capable of explaining the calcium-dependent inhibition of ADP-stimulated respiration . The model used in this study includes mechanistic descriptions of mitochondrial metabolism and energetics in addition to calcium uptake , sequestration , and efflux pathways . In doing so , the first set of experiments produced a list of hypotheses which we began to systematically rule out until left with one possible explanation of the data . The results show that oxidative phosphorylation inhibition by calcium overload is due to the formation of calcium phosphate precipitation , not matrix free calcium , inhibiting complex I . This hypothesis is further supported by experimental data showing no inhibition of isolated complex I activity by calcium [21] . Electron micrograph images have localized these phosphate granules in the matrix near or attached to the inner membrane [15 , 49] . A possible explanation of these findings is that calcium phosphate granules act as physical barriers or disrupt complex I activity . Perhaps , complex I is at or near the nucleation site for granule formation and growth . Alternatively , calcium phosphate granules may disrupt cristae structure and impair mitochondrial function . We found no reports suggesting these mechanisms , so they are speculative and remain to be tested . Our isolated mitochondrial results are supported by several in situ findings . For example , mitochondria isolated from whole hearts subjected to ischemia/reperfusion ( IR ) injury contain large quantities of calcium and depressed ATP synthesis rates [50–53] . In addition , a major target of IR injury is complex I inhibition [54–56] . These findings are in direct support of our isolated mitochondrial work . Moreover , mitochondrial calcium content in excess of 10 nmol/mg protein results in the formation of calcium phosphate granules [57] . In our study , mitochondria were loaded with calcium content in the range of 0–500 nmol calcium/mg protein . Thus , our experimental results with isolated mitochondria capture the major features associated with IR injury where complex I activity is inhibited in response to calcium accumulation and calcium phosphate precipitation prior to the onset mitochondrial permeability transition . The dynamic model of mitochondrial bioenergetics developed in this study served as a powerful tool to help analyze the experimental data . With only five adjustable parameters ( two MCU , two NCLX , and a calcium inhibition constant ) , the model is capable of simulating experimental results from two types of respiratory substrates each with four different calcium conditions . And while the model simulations do not pass through all the data points , they faithfully recapitulate the physiological data in a self-consistent , thermodynamically balanced manner . More complex rate expressions for the MCU and NCLX or the inclusion of the sodium-independent calcium efflux pathway [58 , 59] may increase the quantitative agreement between the model and experimental data; however , this would require many more additional parameters and would only marginally improve the model fits . Regardless , the underlying mechanistic information provided by the model would not change . In summary , we show that mitochondrial respiration with complex I substrates during oxidative phosphorylation is inhibited by calcium phosphate accumulation . Respiration with complex II substrates is nearly independent of calcium phosphate accumulation and only shows a very modest decrease in respiration . This decrease in respiration can be attributed to calcium phosphate granules interfering with additional enzymes in addition to complex I . Further experimental work is needed to verify these findings . For example , direct imagining of calcium phosphate granules and their effect on cristae structure using electron cryomicroscopy with coincident mitochondrial functional data could finally elucidate the mechanism behind calcium overload and oxidative phosphorylation inhibition .
The work presented herein conformed to the National Institutes of Health’s Guild for the Care and Use of Laboratory Animals and was approved by Michigan State University’s Institutional Animal Care and Use Committee . Hartley guinea pigs ( 4–6 weeks ) were anesthetized with 5% isoflurane . Euthanasia was carried out by decapitation . All chemicals and reagents were purchased from Sigma Aldrich unless otherwise stated . Calcium Green 5N was purchased from Molecular Probes . Cardiac mitochondria were isolated from guinea pigs ( 350–450 g ) using differential centrifugation as described in Wollenman et al . [60] . Briefly , isolated mitochondria were suspended at 40 mg/ml in isolation buffer consisting of 200 mM mannitol , 50 mM sucrose , 5 mM dibasic potassium phosphate , 5 mM MOPS , 1 mM EGTA , and 0 . 1% ( w/v ) BSA at pH 7 . 15 . Protein was quantified using the BCA assay and BSA standards . To test the effect of calcium on respiration and membrane potential , 0 . 1 mg/ml mitochondria were suspended in respiration buffer consisting of 130 mM potassium chloride , 5 mM dibasic potassium phosphate , 1 mM magnesium chloride , 20 mM MOPS , 0 . 1% ( w/v ) BSA at pH 7 . 1 at 37°C . Due to chemical impurities , the free calcium concentration in the absence of EGTA in the respiration buffer was 4 . 01 +/- 0 . 43 μM . Sodium pyruvate ( 5 mM ) and L-malate ( 1 mM ) or disodium succinate ( 10 mM ) and complex I inhibitor rotenone ( 0 . 5 μM ) were used to energize the mitochondria . ADP was added as a bolus at a concentration of 500 μM . Respiration was measured using an Oroboros O2k , and membrane potential was measured using an Olis DM245 spectrofluorometer and 0 . 1 μM of the lipophilic cationic dye , TMRM . The ratiometric approach ( 546/573 nm excitation , 590 nm emission ) by Scaduto et al . [61] was used . TMRM fluorescent data were normalized to the maximum value obtained using alamethicin ( 25 μg/ml ) and the protonophore , carbonyl cyanide m-chlorophenyl hydrazine ( 1 μM ) , and the minimum value obtained during the leak respiratory state using oligomycin ( 1 μM ) and nigericin ( 2 μM ) . The signal was then transformed using the following equation: %TMRM ( t ) = 100* ( Rmax−R ( t ) ) / ( Rmax-Rmin ) . Buffer calcium contamination was measured using 1 μM of the fluorescent indicator calcium green 5N ( 503 nm excitation , 532 nm emission ) . The measured KD in our conditions was 30 μM obtained by titrating free calcium concentrations using total CaCl2 and EGTA concentrations given by the MaxChelator program [62] . The experimental protocol was as follows: at 0 minutes mitochondria plus 5mM sodium pyruvate and 1 mM L-malate were added; at 5 minutes , a bolus of CaCl2 ( 0 μM , 25 μM and 50 μM ) was given; at 10 minutes , 500 μM of ADP was added; the respiratory rate was followed for an additional 10 to 20 min . When EGTA was present , no calcium bolus was added . The protocol for measuring membrane potential using TMRM was identical , except that the measurements were carried out in a cuvette open to atmosphere . In other respiratory trials , 10 mM disodium succinate and 0 . 5 μM rotenone was added with the mitochondria . In these trials , the boluses of CaCl2 were added at 2 . 5 mins . The 500 μM ADP bolus was added at 5 min . Calpain I , II , and III inhibitors and the calpeptin inhibitor were purchased from Sigma . The calpain 10 inhibitor was synthesized using the Sigma custom peptide synthesis service . The peptide sequence ( CYGAK ) was identical to the sequence reported in Rasbach et al . [33] . Calpain inhibitors ( 10 μM ) were incubated in the presence of energized mitochondria for five minutes before the addition of the CaCl2 boluses . Five minutes after the CaCl2 boluses , a 500 μM bolus of ADP was added . Data are presented as mean +/- standard deviation or 95% confidence intervals , as noted . Data were analyzed and plotted using MATLAB . Confidence intervals were computed using t and chi-square distributions for uncensored samples using an exact method . Data were checked and confirmed for normality using the Shapiro-Wilk test . Statistical significance was tested using an unbalanced ANOVA followed by a post-hoc Tukey’s range test . The n values ranged from 3 to 20 for each condition tested . The computational model of cardiac mitochondrial energetics from Bazil et al . [38] was used as a framework and updated to include additional reactions and rate equations for sodium and calcium homeostasis . A new lumped TCA cycle flux expression was developed , and a new model of complex III kinetics [45] was incorporated into the mitochondrial bioenergetics model . Some sodium and calcium flux expressions and dissociation constants from Bazil et al . [40] or Bazil et al . [46] were also added to the model . Other minor changes were made to the model and discussed further in the Results and Discussion section and given in the Supporting Information ( S1 File ) . Model codes are available in the Supporting Information ( S2 File ) . The DAE’s of the model were numerically integrated using MATLAB ( R2017b ) and the stiff ode solver ode15s using a relative error tolerance of 10−4 and an absolute error tolerance and 10−9 . The implicit method using backward differentiation formulas was used to solve the system of DAEs . Parameter optimization was done using a desktop PC ( 64-bit operating system and x64-based processor Intel® core ™ i7-7700 CPU @3 . 60GHz and 16 GB RAM ) using the Parallel Computing Toolbox . Manual determination of parameter upper and lower bounds was followed by the gradient-based optimization algorithm , fmincon . | Mitochondrial calcium handling has been studied for nearly half a century . As we understand it today , low concentrations ( 1–10 nmol/mg mitochondria ) of calcium exert beneficial effects on energy transduction . And high concentrations ( >500 nmol/mg mitochondria ) lead to respiratory uncoupling and membrane damage . But relatively little is known about the effect of moderate calcium concentrations ( 10–500 nmol/mg mitochondria ) on mitochondrial function . At these concentrations , mitochondrial membrane integrity remains intact and energized , while ATP synthesis becomes significantly impaired . Prior studies have postulated several possible mechanisms , but the precise consequence of calcium overload on mitochondrial ATP production remained obscure . In this study , we combine experimental and computational approaches to show that calcium phosphate precipitation , as opposed to matrix free calcium , inhibits respiratory function at complex I just enough to limit proton pumping during oxidative phosphorylation and decrease ATP synthesis rates . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"protons",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"phosphates",
"metabolic",
"processes",
"membrane",
"potential",
"permeability",
"electrophysiology",
"simulation",
"and",
"modeling",
"oxidative",
"phosphorylation",
"materials",
"science",
"mitoc... | 2019 | Calcium phosphate precipitation inhibits mitochondrial energy metabolism |
Hendra and Nipah viruses ( genus Henipavirus , family Paramyxoviridae ) are highly pathogenic bat-borne viruses . The need for high biocontainment when studying henipaviruses has hindered the development of therapeutics and knowledge of the viral infection cycle . We have performed a genome-wide siRNA screen at biosafety level 4 that identified 585 human proteins required for henipavirus infection . The host protein with the largest impact was fibrillarin , a nucleolar methyltransferase that was also required by measles , mumps and respiratory syncytial viruses for infection . While not required for cell entry , henipavirus RNA and protein syntheses were greatly impaired in cells lacking fibrillarin , indicating a crucial role in the RNA replication phase of infection . During infection , the Hendra virus matrix protein co-localized with fibrillarin in cell nucleoli , and co-associated as a complex in pulldown studies , while its nuclear import was unaffected in fibrillarin-depleted cells . Mutagenesis studies showed that the methyltransferase activity of fibrillarin was required for henipavirus infection , suggesting that this enzyme could be targeted therapeutically to combat henipavirus infections .
Viruses from the family Paramyxoviridae are highly important pathogens impacting human health , collectively causing hundreds of thousands of deaths globally each year [1] . Respiratory syncytial virus ( RSV ) , human metapneumovirus and parainfluenza viruses cause respiratory disease in infants , young children and the elderly . Furthermore , a marked increase in outbreaks of the highly pathogenic Hendra virus ( HeV ) has recently been observed ( 35 of 51 reported outbreaks have occurred since 2011 ) [2] , while annual outbreaks of Nipah virus ( NiV ) in Bangladesh result in 70–100% mortality [3] . Despite the increasing incidence of infections , there are limited vaccines or therapies to treat infections in humans , and many aspects of paramyxovirus pathogenesis remain poorly understood . Such a major global public health problem calls for the development of novel antiviral strategies . The HeV and NiV genomes ( henipavirus genus ) consist of six genes encoding nine proteins , and the virus exploits host cellular processes to complete its infection cycle . Both viruses enter cells via the ephrin-B2 and/or ephrin-B3 receptors [4] . The reliance on host gene products for virus infection can be determined at a genome-wide level using high-throughput RNA interference ( RNAi ) screening strategies . There have been many full or partial-genome RNAi screens of host-virus interactions , including orthomyxoviruses [5 , 6 , 7 , 8 , 9] , retroviruses [10 , 11 , 12] and flaviviruses [13 , 14] . To date there has not been a genome-wide siRNA screen of host genes required for paramyxovirus infection . Genome-wide siRNAs screening involves the use of robotics and advanced imaging equipment , as such there are technical and practical challenges to performing high-throughput screens at biosafety level ( BSL ) -4 , the highest level of physical containment . Thus , previously published genome-wide screens for BSL-4 viruses have used surrogate viruses , such as pseudotyped particles , and have been performed under BSL-2 conditions [15 , 16] . These surrogate viruses , which often encode only the fusion glycoproteins of the viruses being studied , might not recapitulate the complex and dynamic interactions of all viral genes and proteins acting concurrently in the host cell during a live virus infection . Screen hits from these studies are also often limited to host factors required for early steps of the virus life cycle . Recent outbreaks of BSL-4 agents such as Ebola virus highlight the need for advanced technological and research capabilities , such as high-throughput experimentation with live viruses at BSL-4 , to develop novel intervention strategies . In this study we have performed a genome-wide SMARTpool siRNA screen of protein-coding genes required for HeV infection . To our knowledge , this is the first such study conducted for a paramyxovirus and for a live BSL-4 pathogen . We demonstrate that the henipaviruses require an overlapping subset of host gene products for infection , and that early steps of replication are critically dependent on the nucleolar methyltransferase fibrillarin .
To facilitate high-throughput screening , we utilized a fully infectious recombinant HeV expressing a luciferase reporter construct [17] . An siRNA library targeting 18 , 120 protein-coding genes in arrayed format was transfected into HeLa cells that supported both HeV infection and siRNA transfection . After 72 h of target knockdown , cells were infected with recombinant HeV in a BSL-4 laboratory , with luciferase readings taken 24 h post-infection . Cell viability was measured at the time of infection in parallel plates . Positive and negative controls were used to evaluate transfection efficiency and assay readout robustness on a per-plate basis . As a negative control , cells were transfected with a SMARTpool siRNA ( siNEG ) that does not target any gene and that did not impact HeLa cell numbers or HeV infection compared to mock ( lipid only transfected ) cells ( Fig 1A ) . As a positive control for reducing HeV growth , cells were transfected with an siRNA targeting firefly luciferase ( siLUC ) that inhibited luciferase reporter expression without impacting cell numbers . Cells were also transfected with a “death control”–a SMARTpool siRNA targeting polo-like kinase 1 ( PLK1 ) , a gene associated with apoptosis induction [18] . Death of cells following depletion of PLK1 provided a transfection control in addition to an indirect positive control for inhibition of HeV growth . Applying a robust Z score normalisation across all screen plates , a typical hit-identification strategy for siRNA screens [19 , 20] , we identified 585 and 630 genes , respectively , that statistically promoted or suppressed HeV infection without adversely impacting cell numbers ( Fig 1B , S1 Table and S2 Table , respectively ) . At the completion of the primary SMARTpool screen , we selected 200 proviral genes based on rank for a secondary screen . Validation was performed by deconvoluting the siRNA pools into the four constituent duplex siRNAs and screening each individually using the same assay format . By this measure , we identified 20 high- and 46 medium-confidence genes ( >2 standard deviations from mean mock values for 4/4 or 3/4 , or 2/4 siRNAs , respectively ) required for HeV infection ( S3 Table ) . We also identified 78 low-confidence ( 1/4 siRNAs ) genes ( S3 Table ) . The sub-cellular localization of the 66 high- and medium-confidence genes is shown in Fig 1C . A noticeable observation is the large proportion of genes , including seven out of the 10 with the greatest impact on virus infection , associated with the nuclear or nucleolar compartments . We next examined whether the 66 high- and medium-confidence genes required for infection by the reporter HeV were also required by wild-type HeV and the related NiV . Following siRNA-mediated knockdown of candidate genes , HeLa cells were infected with wild-type HeV or NiV ( MOI 0 . 01 for 48 h ) , and infectious virus titres were measured by TCID50 assays . RNAi-mediated silencing of 43 of the 66 genes resulted in significant decreases in HeV titres ( S4 Table ) . The greatest impact was observed from silencing fibrillarin ( FBL ) , which resulted in an approximate 99 . 9% reduction in virus titres . The majority ( 41 of 43 ) of genes required for HeV infection were also required for NiV infection ( Fig 1D and S4 Table ) . Collectively , these data demonstrate that HeV and NiV exploit similar host gene products for infection as well as validating the recombinant virus used in the genome-wide RNAi screen as a faithful reporter of wild-type HeV infection . To verify target knockdown , we measured mRNA levels of 10 host genes , which , according to TCID50 analysis , most impacted HeV infection . For all 10 genes , siRNAs targeting these genes caused >80% reduction in gene mRNA expression ( Fig 1E ) . The requirement of the nucleolar protein FBL for henipavirus infection was intriguing given that , like most negative-stranded RNA viruses , henipaviruses replicate in the cell cytoplasm [21] . As both HeV and NiV infection was inhibited more than 99 . 9% by FBL knockdown ( S4 Table ) , we investigated the role of FBL in HeV infection in greater detail . We first validated the knockdown of FBL at the protein level . Cell lysates collected from HeLa cells transfected with siNEG showed a single band corresponding to FBL ( ~39 kDa ) by Western blotting ( Fig 2A ) . In contrast , lysates from cells transfected with the SMARTpool siFBL siRNA used in the genome-wide screen , or the constituent siRNA duplexes , showed no detectable FBL protein at 72 h post transfection . A second ON-TARGETplus SMARTpool of siFBL siRNAs with distinct nucleoside sequences featuring chemical modification to reduce off-target effects [22] also depleted FBL ( Fig 2A , siFBL ( 2nd pool ) ) . We next assessed the impact of FBL knockdown on the production of infectious wild-type HeV . Using the SMARTpool siFBL , or the individual duplexes that comprised the pool , a significant reduction ( >99% ) in HeV titers was observed in supernatant collected 48 h after infection ( Fig 2B ) . The ON-TARGETplus SMARTpool also inhibited HeV infection , providing further confidence that HeV infection was inhibited by the loss of FBL . We next assessed the impact of FBL knockdown on cell health . Transfecting HeLa cells with individual duplexes siFBL-1 , -2 and -3 showed a mild impact on cell numbers in conjunction with excellent target knockdown . For siRNA-4 and both SMARTpools , cell numbers were not significantly different to mock or siNEG controls ( Fig 2C ) . An Alamar blue assay showed no significant change in metabolic activity in cells treated with the siFBL SMARTpool for 72 h , compared to cells treated with siNEG ( Fig 2D ) . By contrast , cells transfected with siPLK1 showed significantly reduced metabolic activity . To unambiguously demonstrate that the reduction in HeV infection was mediated by the absence of FBL we conducted a rescue experiment . HeLa cells were transfected with siFBL-2 to knock down endogenous FBL expression , then transfected with a plasmid encoding a FBL gene containing silent mutations in the siFBL-2 target sequence ( pCMV2-FBL ) followed by infection with HeV . Transfecting cells with siFBL-2 decreased the percentage of cells with detectable viral antigen staining by ~70% . Co-transfection of siFBL and pCMV2-FBL successfully restored HeV infection to a level similar to non-transfected cells ( Fig 2E ) . Notably , over-expressing FBL with pCMV2-FBL did not increase HeV infection compared to control cells . This was further confirmed by a separate experiment where pCMV2-FBL over-expression was titrated ( S1 Fig ) . Importantly , silencing FBL using the SMARTpool siRNA depleted FBL protein , inhibited HeV infection but did not negatively impact cell numbers or cell viability . All of the subsequent experiments were therefore performed with siFBL SMARTpool unless indicated . We sought to determine whether other paramyxovirus infections were dependent on FBL . Members of the family Paramyxoviridae are divided into two subfamilies ( Paramyxovirinae and Pneumovirinae ) where HeV and NiV belong to the genus Henipavirus in the subfamily Paramyxovirinae . We opted to test viruses belonging to different genera in the same subfamily: measles virus ( MeV , genus Morbilivirus ) and mumps virus ( MuV , genus Rubulavirus ) , and a virus belonging to the subfamily Pneumovirinae , RSV ( genus Pneumovirus ) . Genome replication for all of these viruses occurs solely in the cytoplasm . In addition , since influenza viruses ( family Orthomyxoviridae ) are known to replicate primarily in the host nucleus [23] , a laboratory strain ( A/WSN/33 ) , which has been used in a number of RNAi screens [5 , 6] , was included in our study for comparison . We observed a significant reduction in virus titres for HeV , NiV , RSV , MeV and MuV but not for A/WSN/33 in cells transfected with siFBL ( Fig 3A ) . To confirm that the reduction in the other paramyxoviruses was caused by FBL depletion , the rescue experiment described earlier ( Fig 2E ) was repeated with RSV . Fig 3B demonstrates that FBL is a factor required for RSV infection . We also used immunofluorescence imaging and software quantification as the readout for virus infection . Cells transfected with siNEG , then infected with HeV for 48 h , showed extensive HeV phosphoprotein ( P ) staining , whereas cells depleted of FBL showed almost no viral protein expression ( Fig 3C ) . Depleting cells of FBL also significantly reduced the proportion of cells stained for RSV nucleoprotein ( Fig 3C ) , however , not to the same extent as HeV or NiV . By contrast , depleting cells of FBL did not alter influenza A/WSN/33 nucleoprotein staining ( Fig 3C ) . FBL is the catalytic component of small nucleolar ribonucleoprotein ( snoRNP ) complex responsible for site-specific 2’O-methylation of ribose found on ribosomal RNAs [24] . Pre-ribosomes undergo chemical modifications such as 2’O-methylation and pseudouridylation as part of their processing during ribosomal biogenesis [24] . As depleting cells of FBL would reasonably be expected to impact ribosome biogenesis and protein translation , and therefore may inhibit virus infections non-specifically , the lack of inhibition of A/WSN/33 infection in FBL-depleted cells demonstrates that FBL is specifically required for paramyxovirus infection . This is also supported by previous RNAi screens that do not link FBL with the infection cycle of several strains of influenza virus [5 , 6 , 7 , 8 , 9] . We sought to determine which stage of the henipavirus infection cycle is dependent on FBL . We first tested the impact of FBL on HeV cell entry using an established cell-cell fusion assay [25] . When incubated with effector cells expressing HeV-F and HeV-G proteins , target cells depleted of FBL exhibited modestly higher fusion levels with effector cells compared to target cells transfected with siNEG ( Fig 4A and 4B ) , suggesting that FBL is not required for HeV cell entry . As a positive control , depleting cells of the HeV entry receptor ephrin-B2 ( using a SMARTpool siRNA , siEFNB2 ) decreased cell-cell fusion by 70% relative to siNEG . In order to investigate whether FBL is required for viral genomic replication and transcription , we first performed a timecourse experiment to characterize the single-cycle replication kinetics of HeV in HeLa cells . A previous study has shown that in HeLa cells , NiV buds from the plasma membrane at or prior to 24 h post-infection ( p . i . ) , concurrent with the earliest timepoint for observable syncytia [26] . Consistent with this report , HeLa cells infected with a high MOI of HeV started producing infectious virions ( above inoculum levels ) between 12 and 24 h p . i . ( Fig 4C ) . This indicates that the length of one cycle of HeV infection in HeLa cells is approximately 12 to 24 hours . Consequently , a validated TaqMan qPCR assay [27] was used to measure intracellular viral RNA levels at 8 , 12 and 24 h p . i . . Intracellular viral RNA levels started increasing above inoculum levels between 8 and 12 h p . i . ( Fig 4D ) , which is consistent with the replication kinetics observed with the virus production data ( Fig 4C ) . More importantly , within this single-cycle infection period ( 12 h p . i . ) , knockdown of either FBL or EFNB2 significantly reduced intracellular viral RNA levels relative to siNEG . Additionally , at 24 h p . i . , virus titers in siFBL and siEFNB2 groups showed significantly reduced levels compared to siNEG ( Fig 4C ) . Finally , the impact of FBL on viral replication was also assessed in a minigenome assay not involving infectious virus . HeV gene expression levels were significantly reduced in HeLa cells transfected with siFBL compared to siNEG ( Fig 4E ) . As expected , viral protein production ( P protein and N protein ) was almost completely abolished in HeV-infected cells treated with siFBL ( Fig 4F and S2 Fig , respectively ) . Collectively , these results indicate that FBL is required for viral RNA synthesis during the pioneering rounds of infection . FBL , along with 18 out of 43 other validated high-confidence hits , localizes primarily to the nucleus or nucleolus of the cell ( S3 and S4 Tables ) . Interestingly , even though henipaviruses replicate in the cytoplasm , HeV matrix protein ( HeV-M ) has been shown to localize mainly in the nucleus early during infection and then throughout the cytoplasm later in infection in mammalian cells [28] . Wang et al . also demonstrated that NiV-M transits through the nucleolus [26] . We investigated whether FBL and HeV-M colocalize in the nucleolus during live virus infection . Using confocal microscopy , we observed the majority of FBL staining in the nucleolus , with weaker staining in the nucleoplasm . Strong HeV-M staining was also found in the nucleolus and localized at the plasma membrane , with weaker staining in the nucleus and cytoplasm ( Fig 5A ) . Since confocal microscopy suggests a co-localization of FBL and HeV-M in the nucleolus of infected cells , we tested whether the proteins form complexes by reciprocal co-immunoprecipitation . Fig 5B shows that regardless of whether it is FBL or HeV-M that was immunoprecipitated from co-transfected cells , the other protein was detected as a co-immunoprecipitant . Control immunoprecipitations using lysates from single-DNA transfections that express either epitope-tagged FBL or HeV-M demonstrate the specificity of the pull-down . These co-immunoprecipitations , while demonstrating that FBL and HeV-M are associated , do not reveal whether the interaction is direct or indirect via a complex involving other cellular components . We next considered the functional relevance of the observed physical association between FBL and HeV-M in the nuclear compartment . The role of henipavirus M in the nucleus is unclear , though Wang et al . showed that its nuclear targeting and chemical modification ( by ubiquitination ) are required prior to its nuclear export and for virion budding during the later stages of viral morphogenesis [26] . The matrix proteins which are targeted into the nucleus during the early stages of henipavirus infection consist of the initial batches of newly synthesized M , and quite likely , the M derived from the original infecting virions ( i . e . pioneer M ) . Since FBL is required for viral RNA synthesis during early steps of henipavirus replication , any functional relevance of FBL with M should apply primarily to M synthesized prior to or at this stage of infection . Hence , we assessed whether FBL is required for nuclear targeting ( import ) of M for this stage of the infection cycle . For these experiments , our strategy was to use cells transfected with a cDNA plasmid encoding the ORF of myc-tagged HeV-M , since viral protein synthesis was extremely limited during HeV infection of FBL-knockdown cells ( Fig 4F and S2 Fig ) . Previous studies and those of our own have also shown that the trafficking kinetics of M from a transfected plasmid recapitulates those of M from live henipavirus infections ( [26] and S3 Fig ) . At 12 h post-transfection with plasmid , which is a time prior to HeV-M nuclear egress in live infection , HeV-M was observed largely in the host cell nucleus and nucleolus in cells transfected with siNEG or siFBL and there was no discernible difference in the nuclear and nucleolar import of M between the two treatment groups ( Fig 5C ) , where M colocalized with the nucleolar marker nucleolin . Moreover , in FBL-depleted cells , HeV-M is able to complete its entire functional life cycle and produce virus-like particles , albeit with a reduction in budding efficiency ( S4 Fig ) . Therefore , FBL associates with HeV-M in a nucleolar complex but is not required for M nuclear translocation early during infection . In its conventional cellular function of pre-rRNA methylation , human FBL forms a snoRNP complex composed of FBL , NOP56 , NOP58 , NHP2L1 and small guide RNA molecules [29] . To evaluate whether the reliance of henipaviruses on FBL is mediated through the role of FBL in pre-rRNA methylation , we used SMARTpool siRNAs to reduce expression of every member of the snoRNP complex and measured the impact on HeV infection . Cells depleted of NHP2L1 were not viable , while cells depleted of NOP56 and NOP58 remained viable ( Fig 6A ) . Depletion of either NOP56 or NOP58 inhibited HeV infection , as measured by TCID50 assays and immunofluorescence ( Fig 6B–6D ) . However , this impact was significantly less compared to FBL knockdown , which is in agreement with our result from the screen where NOP56 and NOP58 showed an impact on HeV infection but below statistical significance ( Z scores -1 . 99 and -1 . 05 , respectively ) . These results suggest that the role of FBL in henipavirus infection is mediated via its role in the methylation of pre-ribosomes . FBL is the enzymatic subunit of the snoRNP complex , and acts by catalyzing the transfer of a methyl donor from a bound cofactor S-adenosyl methionine ( SAM ) to riboses of the target pre-rRNA [29] . To assess whether the 2’O-ribose methyltransferase activity of FBL is required for henipavirus infection , we performed a mutagenesis study where the main residues involved in FBL binding to its methyl donor cofactor SAM [30] were mutated . Based on mutagenesis analysis of the yeast FBL ( NOP1 ) [30] and the crystal structure of the human FBL-MTA ( a chemical analogue of SAM ) complex , two conserved residues critical for FBL catalytic function were selected ( Fig 7A ) . Alanine mutations at yeast residues corresponding to human residues at E191 and D236 result in greatly impaired FBL methylation activity in vitro [30] . These mutants were applied to an assay similar to the rescue experiment ( Fig 2E ) , using a cDNA construct expressing recombinant FBL resistant to RNAi silencing . Next , the catalytic residues were individually substituted to alanine residues and expression of the mutants in HeLa cells were confirmed by Western blotting . Expression levels of D236A were low ( Fig 7B ) , possibly due to improper protein-folding , and so this mutant was not tested further . The E191A mutant protein was expressed ( Fig 7B ) and when tested for its ability to rescue HeV infection , showed impaired virus infection compared to control cells ( Fig 7C ) . On the contrary , the E191A mutant was able to rescue RSV infection ( Fig 7D ) , suggesting that the catalytic activity of FBL is required for henipavirus infection but not for paramyxoviruses from other genera . In addition , co-immunoprecipitation assay indicated that the E191A mutant does not lose binding to HeV-M ( S5 Fig ) . These observations are consistent with the notion that FBL supports henipavirus infection via its role in pre-rRNA processing .
Research into henipaviruses has been limited by multiple factors–not least the fact that few laboratories in the world have appropriate facilities to handle BSL-4 pathogens . Accordingly , many determinants of pathogenesis have remained unclear . Here we present the first high-throughput RNAi screen performed under BSL-4 containment , as well as the findings from a genome-wide analysis of host genes required for live henipavirus infection . The screen resulted in the identification of multiple host genes involved in ribosomal biogenesis ( e . g . RPL and RPS family members , ESF1 ) , nuclear export/import ( e . g . XPO1 , KPNA3 ) and transcriptional regulation ( e . g . BTF3 , SP7 ) that are important for henipavirus infection . Stages of pre-ribosomal processing which are associated with the validated gene hits include rDNA transcription ( POLR3E ) , pre-rRNA cleavage ( components of the U3 small subunit processome DDX10 and IMP4 ) [31] , subunit assembly ( GTPBP4 ) , chemical modifications of pre-rRNA ( FBL , RPL13A ) [32 , 33 , 34] , as well as nuclear export ( XPO1 ) ( S3 Table , S6 Fig ) . In general , proper ribosomal processing is required for canonical ribosome function ( cap-dependent translation ) and for efficient cellular and viral protein translation . However , more recent evidence indicates that chemical modifications of pre-ribosomes , ( e . g . 2’-O-ribose methylation ) are dispensable for downstream canonical ribosome functions [32 , 33 , 34] . Undermethylated or non-methylated ribosomes exhibit normal ribosomal RNA processing , polysome formation , global translational activity , and translational fidelity [33] . Interestingly , FBL is the catalytic subunit directly responsible for 2’-O-ribose methylation of pre-rRNA [30] , and yet it exhibited the highest impact on henipavirus infection of all the proviral hits from the screen . This suggests that FBL is required by henipaviruses for a unique function other than canonical cap-dependent translation . This notion is consistent with our observations that knockdown of FBL does not affect the synthesis of HeV matrix proteins ectopically expressed from a transfected plasmid ( Figs 5 and S4 ) or the infection and protein synthesis of influenza virus ( Fig 3 ) . Several studies have shown that methylation of rRNA , though not affecting traditional ribosomal functions , does regulate the type of mRNA transcripts targeted for translation [32 , 33 , 34] . Specifically , inhibition of rRNA methylation , either by depleting FBL or by drug treatment , inhibits translation of host transcripts with internal ribosome entry site ( IRES ) elements in their 5’ UTR [33 , 35] . Interestingly , another validated high-confidence hit from our screen , ribosomal protein L13a ( RPL13A ) ( S4 Table ) , is also required for rRNA methylation and IRES-mediated translation [33] . Co-immunoprecipitation and colocalization assays further show that FBL associates in a complex with RPL13A and a C/D box snoRNA , U15 [32] . We demonstrate that FBL is not required for henipavirus entry ( Fig 4A and 4B ) but is instead essential for viral RNA synthesis ( Fig 4D and 4E ) . Consistently , in FBL-deficient cells , the most abundant viral proteins , P and N , were minimally detected ( Fig 4F and S2 Fig , respectively ) . It is conceivable that FBL acts as a proviral factor by modulating the translation of host genes involved in viral replication . Many cellular mRNA transcripts that are IRES-dependent encode for stress response genes [36 , 37] , including inhibitors of apoptosis , cell proliferation factors , and immune-suppressive regulators . The role of FBL in modulating the host microenvironment to support infection will be an interesting subject for future studies . We have shown that in order to impact henipavirus infection , FBL requires its methylation activity ( Fig 7 ) . In eukarya FBL can form a complex with NHP2L1 , NOP56 and NOP58 to perform its 2’-O-methyltransferase activity [38] and we have shown that HeV infection is decreased in the absence of any one of the members of this complex ( Fig 6 ) , with the exception of NHP2L1 which impacted cell viability . Murray et al reported that disruption of expression of C/D small nucleolar RNAs expression , which also participate in 2’-O-methylation with FBL , confers resistance to viral infection such as RSV [39] , thus supporting the idea that host-mediated methylation plays a vital role in paramyxovirus infection . Until recently , FBL was believed to solely modify RNAs , but Tessarz et al demonstrated that FBL is responsible for rDNA transcriptional regulation by methylating Q104 of human histone H2A , opening up the possibility that other host or even viral proteins could be methylated by FBL [40] . In addition to FBL , SET and MYND domain containing 2 polypeptide ( SMYD2 ) , also a protein N-methyltransferase , is another validated high-confidence hit ( S4 Table ) . It specifically methylates and dimethylates histone H3 , as well as other non-histone proteins [41 , 42] . This further validates that methylation plays critical roles in henipavirus infection . We have shown that in cells depleted of FBL and infected with HeV , viral proteins are detected at extremely low levels ( Fig 4F and S2 Fig ) . However , in cells expressing FBL we found that FBL and HeV matrix protein ( M ) colocalized in the nucleolus ( Fig 5A ) . Paramyxovirinae M are known to traffic through the nucleus and nucleolus and that this step is required for efficient viral budding and is regulated by ubiquitination [26 , 43] . Although current microscopy techniques do not allow detection of M released from the incoming virions ( pioneer M ) , we observed that cells lacking FBL do not show changes in nuclear import of ectopically expressed HeV-M at early timepoints , which suggests that pioneer M can still enter the nucleus even in the absence of FBL ( Fig 5C ) . Moreover , we confirmed binding of FBL and HeV-M ( Fig 5B ) and although Pentecost et al did not identify FBL as a direct binding partner for Paramyxovirinae M , their protein interactome analyses suggested interactions between M and numerous nucleolar proteins such as NOP58 [26 , 43] which forms a complex with FBL , supporting the idea of a functional interaction between FBL and HeV-M . Moreover , Sun et al demonstrated that the binding between henipaviruses M and nuclear proteins can be disrupted by RNase A treatment , suggesting an indirect binding through cellular RNAs [44] . Interestingly , 19 out of 43 top validated hits have a nuclear or nucleolar localization , emphasizing the importance of nuclear proteins in HeV infection . It has recently been shown that HeV-M and NiV-M bind to the nuclear protein acidic leucine-rich nuclear phosphoprotein 32 family member B ( ANP32B ) , which is involved in nuclear mRNA export processes , regulation of gene expression and apoptosis , and that this binding increases nuclear retention of M [45] . Our screen demonstrated that RNAi-mediated silencing of ANP32B increases ( but not significantly ) virus infection ( Z score 1 . 54 ) , suggesting that ANP32B may have a hitherto undescribed antiviral role . Paramyxovirinae M main role in viral infection is viral budding and we have shown that viral budding still occurs even in the absence of FBL even though its efficiency was decreased , which suggests that at later times in HeV infection FBL could play a role in M trafficking and viral budding ( S4 Fig ) . Matrix proteins have also been shown to be involved in regulation of viral transcription [43 , 46 , 47] , RNA binding [47] and modulation of host transcription [48] . It could be speculated that HeV-M traffics to the nucleolus and interacts with FBL to regulate host gene expression to facilitate viral replication . Alternatively , HeV-M could impact FBL activity such as enhancing its methylation activity . Future work will aim to understand whether FBL facilitates the infection cycle of paramyxoviruses differentially by more than one mechanism . In this regard , several distinctions between viruses are evident from our study and others: ( i ) the impact of FBL on virus titres and/or virus protein production were greater for the henipaviruses than MeV , MuV and RSV ( Fig 3 ) ; ( ii ) the catalytic activity of FBL was required for HeV infection but not RSV infection ( Fig 7 ) , and ( iii ) while HeV-M traffics to the nucleolus and associates with FBL ( Fig 5 ) , RSV-M does not traffic to the nucleolus [48] , despite RSV infection being impacted by FBL ( Fig 3 ) . In summary , this screen dataset furthers the understanding of the infection cycle of henipaviruses and is a resource for the study of related paramyxoviruses–a virus family that impacts both human and animal health . This study reveals a previously unappreciated role for nucleolar proteins with methyltransferase activity such as FBL in henipavirus infection , and suggests that methyltransferase enzymes represent a potential target for development of an anti-henipavirus drug . This work also serves as a blueprint for how high throughput RNAi screens can be performed under high biocontainment conditions .
HeLa cells ( ATCC CCL-2 ) , Hep-2 cells ( ATCC CCL-23 ) , African green monkey kidney epithelial Vero cells ( ATCC CRL-81 ) , Madin–Darby Canine Kidney ( MDCK ) cells ( ATCC CCL-34 ) and HEK 293T cells ( ATCC CRL-3216 ) were maintained as described previously [49 , 50] . All virology work was conducted at the CSIRO Australian Animal Health Laboratory . Recombinant HeV , wild type HeV ( both Hendra virus/horse/1994/Hendra ) , NiV ( Nipah virus/Malaysia/human/99 ) , MeV ( wild type Edmonston strain ) , MuV ( Enders strain ) and RSV ( strain A2 ) were passaged in Vero cells . Influenza A/WSN/33 ( H1N1 ) ( kind gift , Professor Lorena Brown , University of Melbourne ) was passaged in the allantoic fluid of 10-day embryonated specific pathogen-free chicken eggs ( Australian SPF Services , Cadello , VIC , Australia ) . All viruses were aliquoted and stored at −80°C for inoculations . HeLa cells ( 1100 cells/well ) were transfected in 384-well plates ( Corning product # 353988 ) with siGENOME SMARTpool siRNAs ( final concentration 40 nM ) using DharmaFECT ( DF ) 1 lipid transfection reagent ( 0 . 03 μL/well ) ( all reagents from Dharmacon RNAi Technologies , GE , USA ) . The siCONTROL Nontargeting siRNA #1 ( catalog # D-001210-01-05 , referred to here as “siNEG” ) and a siRNA targeting firefly luciferase ( #P-002099-01-20 ) acted as negative and positive controls respectively . Mock ( transfection lipid only ) wells were also included . Genome-wide siRNA libraries ( catalog numbers in S1 and S2 Tables ) were screened at the Victorian Centre for Functional Genomics ( VCFG ) . Each well in the siGENOME SMARTpool siRNA library contained 4 distinct siRNAs targeting different sequences of the target transcript . Deconvolution validation screens were performed using individual siRNA duplexes , all 4 on the same library plate ( catalog numbers in S3 Table ) at 25 nM with assay conditions as above . Cells were reverse transfected using the Sciclone ALH3000 ( Caliper Life Sciences , Hopkinton , MA ) and BioTek 406 ( BioTek , Winooski , VT ) liquid handling robotics . Cells were transfected in quadruplicate and divided into 2 groups of duplicate plates for nuclei quantitation and duplicate plates for HeV infection . At 72 hours post transfection , in parallel with the point of HeV infection , cell viability for each well was assessed by fixing ( 4% paraformaldehyde for 10 min ) and staining plates with the nuclear stain 4' , 6-Diamidino-2-Phenylindole , Dihydrochloride ( DAPI ) ( Invitrogen , Carlsbad , CA; 1 μg/ml for 20 min in PBS ) . The number of cell nuclei per well was quantitated for 25 fields using a 20 × objective on a Cellomics Arrayscan VTi microscope using the Target Activation bioapplication of the Cellomics Scan software ( iDev workflow ) . ( Thermo Fisher , Waltham , MA ) . HeV growth was measured in separate plates suitable for luminescence assays ( Corning product # 353988 ) . 72 h post-transfection , cells were infected with recombinant HeV ( MOI 0 . 1 using a BioTek 406 liquid handler housed in a class II biosafety cabinet at BSL-4 . At 24 hours post-infection , media was removed and 20 μL of PBS added per well . Luminescence was then measured by addition of 20 μL of Bright-Glo Luciferase reagent ( Promega , Madison , WI ) and reading on a Synergy H4 multimode microplate reader ( BioTek ) . Data analysis was automated using a custom R script which combined and analysed the luciferase ( HeV growth ) and DAPI ( cell viability ) raw data files to generate a summarized report spreadsheet with raw and normalised values . Both HeV infection and cell viability were normalized to the average of the mock control readout per plate . siRNA transfections that resulted in >50% average reduction in cell viability compared to mock controls were scored as toxic ( LC , low cell count ) and excluded from further analyses . The experimental robustness was evaluated for each screened plate using the Z’ factor calculation [19] , comparing the negative control ( siNEG ) , positive control ( siLUC ) and death control ( siPLK1 ) for both cell viability and HeV infection . Robust z-scores utilising the median and median absolute deviation ( MAD ) of all control-normalised sample values were generated across all sample wells and averaged per duplicate plate pair . Robust z score = ( sample value-sample median ) /sample median absolute deviation were used as the bio-identification method [19 , 51] . For experiments following the genome-wide screen , HeLa cells ( 9000 cells/well ) were reverse transfected in 96 well plates with siFBL , siNEG or siEFNB2 siGENOME SMARTpool siRNAs ( final concentration 40 nM ) using DF1 ( 0 . 12 μL/well ) for 72 h . Target gene knockdown was assessed 48 h post transfection with siRNAs using DF1 according to manufacturer’s instructions . RNA purification , cDNA synthesis and quantitative real-time PCR were performed as described previously [49] . HeV RNA was quantified using a HeV Taqman one-step PCR assay described previously [27] . Protein lysates were collected from cells seeded in 24 well plates , 48 h post-transfection with siRNAs or 24 h post-transfection with plasmid DNA ( 500 ng/well ) using Lipofectamine 2000 . Western blot analyses were carried out according to standard protocols with primary antibodies ( 1:1000 ) against FBL ( monoclonal ( 38F3 ) , Abcam , Cambridge , UK ) , β–actin ( monoclonal ( AC-74 ) , Sigma ) or anti-FLAG ( CSIRO Manufacturing Flagship ) and species-appropriate horseradish peroxidase-conjugated secondary antibody ( 1:10 , 000 dilution ) . Western blot band intensities were quantified using Image Lab software ( Bio-Rad ) . Assays were performed as described previously [49] . Samples were titrated in triplicate in 96-well plates , co-cultured with Vero cells for three days ( HeV , NiV ) or seven days ( MeV , MuV , A/WSN/33 ) , and assayed for cytopathic effect . The infectious titre was calculated by the method of Reed and Muench [52] . Plaque assays were performed using a modified protocol [50] . Briefly , Hep-2 cells were seeded in 24-well plates at 1 . 9 × 105 cells per well . 50 μL of cell supernatants were added to 150 μL of DMEM , and incubated with HEp-2 cells for 4 h , with shaking every 30 min . Inoculum was removed and replaced by overlay ( DMEM/0 . 3% agarose/2% FCS ) . Cells were incubated for 7 days at 37°C 5% CO2 . On day 7 , 2 mL of 1% formaldehyde made up in 150 mM NaCl was added , and left to penetrate overnight . Agarose was removed , and 2 . 5 mL of 0 . 05% neutral red was added . Wells were stained for 1 h , washed and plaques counted . Viable nuclei were quantified using the CellInsight Personal Cell Imager ( Thermo Scientific ) at a magnification of 10 × , with images captured of all areas of each well , as described [49] . Alamar blue ( Invitrogen ) assays were performed following manufacturer’s guidelines . Briefly , cells in a 96 well plate were incubated with 10% ( v/v ) Alamar blue for 4 h at 37°C , followed by fluorescence measured at 570 nm/585 nm excitation emission filter using a Synergy microplate reader ( BioTek ) . Cells were stained to detect HeV/NiV-P , HeV-N , HeV-M or NiV-P as described [28] . Cells were stained to detect RSV nucleoprotein ( N; mouse monoclonal ( 8B10 ) , Sapphire Biosciences , used at 1/1000 ) , influenza virus nucleoprotein ( NP; mouse monoclonal ( AA5H ) , AbD Serotec , 1/500 ) , FBL ( rabbit polyclonal ( Ab5821 ) , Abcam , 1/2000 ) , c-myc ( mouse monoclonal ( clone 9E10 ) CSIRO Manufacturing Flagship , 1ug/ml ) or nucleolin ( rabbit polyclonal ( Ab22758 ) , Abcam , 1/1000 ) . Cells were stained with 1/200 dilution of the appropriate secondary antibody purchased from Invitrogen ( HeV-M , influenza N , RSV N , c-myc: anti-mouse AF488 , HeV/NiV-P , HeV-N , FBL: anti-rabbit AF488 , FBL , Nucleolin: anti-rabbit AF568 ) . Nuclei were counter-stained with DAPI . Confocal images were acquired using a Leica-microsystems TCS SP5 microscope . HeLa cells in 96 well plates were imaged using the CellInsight at a magnification of 10 x , 49 fields/well representing the entire well . The percentage of infected cells was quantified using the Target Activation bioapplication of the Cellomics Scan software and was determined by dividing the number of antigen-positive cells by the total cell number , multiplied by 100 . Fusion between F and G-expressing effector cells and permissive target cells was measured using an assay adapted from [25] . Plasmids encoding HeV-F and HeV-G or vector only control were transfected into HEK 293T ( effector ) cells seeded in 6-well plates ( 2 . 5 μg DNA/well , Lipofectamine 2000 ) and allowed to express for 10 h . HeLa ( target ) cells were seeded in 96 well plates and transfected with siRNAs ( 40 nM ) for 72 h ( as above ) and incubated with the Vibrant DiO cell labelling system for 10 min ( Life Technologies ) to label target cells . Effector and target cell mixtures were incubated in 96-well plates at 37°C for 24 h . The ratio of effector to target cells was 32:1 . Cells were fixed with 4% PFA for 30 min , washed with PBS , and stained to detect HeV-G ( primary antibody 1:1000 anti-G mouse #185 ( 21B6 ) , secondary antibody ( 1:250 anti-mouse Alexa Fluor 568 ) . Nuclei were labeled with DAPI as described above . Cell fusion was measured using the Target Activation bioapplication of the Cellomics Scan software ( iDev workflow ) on the CellInsight . HeLa cells seeded in 24 well plates ( 50 , 000 cells/well ) were transfected with siRNAs ( final concentration 40 nM ) using DF1 as described above . 72 h post-transfection , cells were infected with HeV ( MOI 5 ) for 45 min . After this time , cell media was replaced and following 0 h ( inoculum ) 6 , 8 , 12 and 24 h of infection , cell media was collected to measure virion production by TCID50 assay and cell lysates collected to detect intracellular viral RNA levels by qRT-PCR . A Nanoluciferase-based HeV minigenome assay was adapted from a previously developed NiV minigenome assay [53] . Briefly , a bacteriophage T7 polymerase-based HeV minigenome was synthesized ( Genscript ) expressing a reporter fusion construct of Nanoluciferase ( Promega ) and mNeonGreen fluorescent protein [54] . The open reading frame encoding the reporter fusion protein was flanked by a T7 promoter , hammerhead ribozyme , and HeV leader and N gene at the 3’ end . HeLa cells ( 5× 103 per/well were seeded in 96-well plates . The next day , cells were transfected with siRNAs as described above . 48 h post-transfection of siRNAs , appropriate volumes of HeV support plasmids ( N ( 50 ng/well ) , P ( 32 ng ) and L ( 50 ng ) ) , HeV minigenome ( 120 ng ) and T7pol ( 80 ng ) were prepared in RNase-free TE buffer . For negative controls , the L plasmid was substituted with an equivalent amount of pcDNA 3 . 1 plasmid expressing the red fluorescent protein mCherry ( Clontech ) . Plasmids were mixed with 0 . 6 μL/well LT-1 transfection reagent ( Mirus Bio , Madison , WI ) and 10 μL Opti-MEM/well . Complexes were incubated for 30 min at room temperature before adding to cells . 48 h post-transfection of minigenome plasmids , 50 μL of Nanoluciferase assay buffer solution ( Promega ) was added directly to each well . Well contents were transferred to white plates , and after three minutes , luminescence was read on a plate reader ( HT-Synergy , Biotek ) . Cell Titer Glo 2 . 0 reagent ( Promega ) was added immediately following reporter minigenome luminescence reading to measure cell viability according to manufacturer’s guidelines . The average raw luminescence value for siNEG transfected cells ( n = 16 ) was set as 100% cell viability . Percent ( % ) cell viability for individual wells were calculated by dividing their respective raw luminescence values by the average raw luminescence value for siNEG transfected cells . Reporter nanoluciferase activity for each well was then normalised by dividing the nanoluciferase luminescence value by its respective % cell viability as determined above . HEK293T cells cultured in 6-well plates were transfected with 1 μg of pCMV:FLAG-tagged FBL or pCAGGS:myc-tagged HeV-M or both using Fugene6 transfection reagent ( Promega ) . Cells were harvested 48 h post transfection , washed once with 0 . 5 mL ice-cold PBS and lysed using 450 μL SU buffer ( 150 mM NaCl , 10% ( v/v ) glycerol , 1% ( v/v ) Triton X-100 , 1 mM EGTA , 1 . 5 mM MgCl2 , 20 mM HEPES pH 7 . 5 ) , supplemented with protease inhibitors ( Roche ) and 10 mM EDTA for 20 min on ice . The lysates were centrifuged at 16000 × g for 40 min at 4°C and the SU-soluble fractions were recovered . 180 μL aliquots of rec-Protein-A Sepharose 4B ( Zymed ) were treated with 3 μL of either M2 anti-FLAG Mab ( 4 . 2 μg ) or 3 μL 9B11 anti-myc Mab ( Cell Signaling ) , incubated at 4°C for 1 h and washed once with 0 . 5 mL SU buffer to eliminate unbound Mabs . Aliquots of antibody-bound slurry equivalent to 30 μL of original Protein-A Sepharose were resuspended with 200 μL aliquots of lysate for immunoprecipitation for 2 h at 4°C . The slurry was washed three times with SU buffer and resuspended in 1 × reducing PAGE buffer supplemented with 3 M urea . The entire ORF of human FBL was amplified by PCR from a cDNA library and cloned into the expression vector pCMV2 , generating the pCMV2-FBL plasmid . Site-directed mutagenesis was then performed on pCMV2-FBL using QuikChange Lightning Mutagenesis kit ( Agilent Technologies ) . The resultant plasmid , pCMV2-FBLsi2 , has five silent mutations in the target region of a siFBL-2 . The forward and reverse primers used for generating pCMV2-FBLsi2 were 5’–gttggtcctcttcttggccagattgattaagtcacggccagagcggtggg– 3’ and 5’–cccaccgctctggccgtgacttaatcaatctggccaagaagaggaccaac– 3’ , respectively . Using pCMV2-FBLsi2 as the template , the codons for E191 and D236 critical for catalytic activity were substituted with codons for alanine . Primers used for PCR-based mutagenesis of pCMV2-FBLsi2 were: E191A-F: tagtctatgcagtcgcgttctcccaccgctc , R: gagcggtgggagaacgcgactgcatagacta , D236A-F: gatgtgatctttgctgctgtggcccagccagac , R: gtctggctgggccacagcagcaaagatcacatc . Final sequences of FBL for all plasmids made were confirmed by DNA sequencing . HeLa cells seeded in 96 well plates were transfected with either siNEG , siFBL-2 , or siFBL SMARTpool . At 24 h post siRNA transfection , the cells were transfected with either 125 ng of pCMV2-FBLsi2 , pCMV2-FBLsi2-T173A , pCMV2-FBLsi2-E191A , pCMV2-FBLsi2-F192A , pCMV2-FBLsi2-D236A , or vector only . At 48 h post DNA transfection , the cells were infected with HeV or RSV ( MOI 0 . 1 ) . Inoculum was removed after 1 h , and infection was allowed to proceed for 48 h . The percentage of infected cells was quantified as described above . The statistical analysis of the screen itself is described above . For all other work , the difference between two groups was analyzed by a two-tailed Student's t test and between multiple groups by one-way ANOVA . A P value of <0 . 05 was considered significant . All data points are the average of triplicates , with error bars representing standard deviations . All data are representative of results from at least 2 separate experiments . | The henipaviruses Hendra and Nipah are bat-borne paramyxoviruses that are highly pathogenic in humans . The need for high biocontainment when studying Hendra and Nipah virus biology has hindered the development of therapeutics and knowledge of the viral infection cycle . This study describes a genome-wide functional genomics screen of human host genes required for henipavirus infection , to our knowledge the first such study conducted at biosafety level 4 . Our study demonstrates that henipavirus infection is critically reliant on fibrillarin , a methyltransferase enzyme residing in the cell nucleolus . Despite henipavirus genome replication occurring in the cytoplasm of infected cells , viral RNA synthesis was greatly impaired in cells lacking fibrillarin . Furthermore during the early stages of infection the Hendra virus matrix protein shuttles to the nucleolus and binds fibrillarin . Collectively these results suggest a hitherto unappreciated role for nucleolar host-virus interactions in the early replication phase of henipavirus infection . Finally , mutating the catalytic activity of fibrillarin inhibits henipavirus infection , suggesting that this enzyme could be targeted therapeutically to combat henipavirus infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"transfection",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"respiratory",
"infections",
"hela",
"cells",
"gene",
"regulation",
"henipavirus",
"pathogens",
"biological",
"cultures",
"microbiology",
"pulmonology",
"viruses",
"rna"... | 2016 | Genome-wide siRNA Screening at Biosafety Level 4 Reveals a Crucial Role for Fibrillarin in Henipavirus Infection |
Expression of ABO and Lewis histo-blood group antigens by the gastrointestinal epithelium is governed by an α-1 , 2-fucosyltransferase enzyme encoded by the Fut2 gene . Alterations in mucin glycosylation have been associated with susceptibility to various bacterial and viral infections . Salmonella enterica serovar Typhimurium is a food-borne pathogen and a major cause of gastroenteritis . In order to determine the role of Fut2-dependent glycans in Salmonella-triggered intestinal inflammation , Fut2+/+ and Fut2-/- mice were orally infected with S . Typhimurium and bacterial colonization and intestinal inflammation were analyzed . Bacterial load in the intestine of Fut2-/- mice was significantly lower compared to Fut2+/+ mice . Analysis of histopathological changes revealed significantly lower levels of intestinal inflammation in Fut2-/- mice compared to Fut2+/+ mice and measurement of lipocalin-2 level in feces corroborated histopathological findings . Salmonella express fimbriae that assist in adherence of bacteria to host cells thereby facilitating their invasion . The std fimbrial operon of S . Typhimurium encodes the π-class Std fimbriae which bind terminal α ( 1 , 2 ) -fucose residues . An isogenic mutant of S . Typhimurium lacking Std fimbriae colonized Fut2+/+ and Fut2-/- mice to similar levels and resulted in similar intestinal inflammation . In vitro adhesion assays revealed that bacteria possessing Std fimbriae adhered significantly more to fucosylated cell lines or primary epithelial cells in comparison to cells lacking α ( 1 , 2 ) -fucose . Overall , these results indicate that Salmonella-triggered intestinal inflammation and colonization are dependent on Std-fucose interaction .
Glycosylation is an important type of post-translational modification of proteins and lipids and is involved in the regulation of a wide range of processes at the cellular and molecular level . The gastrointestinal tract is home to a vast array of glycan structures and glycoconjugates [1] , where the mucosal surface is the site of complex interactions between the intestinal microbiota , intestinal barrier , and immune system . The mucosal surface is characterized by a heavily glycosylated mucus layer produced by goblet cells as well as membrane-bound glycosylated proteins and lipids that form them [2] . These glycoconjugates can be utilized by intestinal commensal bacteria and pathogens as molecular attachment sites or as nutrients [3] . Importantly , host-derived glycans can help foster beneficial relationships with symbiotic microbes , such as Bacteroides thetaiotaomicron , by providing an energy source in the absence of dietary polysaccharides [4] . The FUT2 gene encodes the α-1 , 2-fucosyltransferase , a glycosyltransferase well known for its role in the expression of ABH and Lewis histo-blood group antigens on the gastrointestinal epithelium and in bodily secretions . Individuals expressing a functional allele are commonly described as ‘secretors’ whereas those homozygous for loss-of-function mutations display a ‘non-secretor’ phenotype . Variation in host glycosylation may directly influence susceptibility to enteric pathogens such as enterotoxigenic Escherichia coli [5] , Helicobacter pylori [6] , and norovirus [7] . Recent studies have shown the importance of host glycans in supporting a beneficial relationship with the endogenous microbiota by nourishing the microbiota during the stress of systemic infection [8] or by controlling opportunistic pathogens within the microbiota in the context of infection ( e . g . Enterococcus faecalis ) [9] . The non-secretor phenotype is also associated with an increased risk to develop chronic inflammatory bowel diseases [10] . This is possibly due to the altered composition of the intestinal microbiota , which may in turn influence the capacity of pathogenic bacteria to bind to host mucosal surface structures [11] . Salmonella enterica serovar Typhimurium ( S . Typhimurium ) is one of the most successful mucosal pathogens , colonizing the human gastrointestinal tract and causing severe inflammatory diarrhea [12] . S . Typhimurium carries several virulence genes including fimbrial adhesins , which are hair-like appendages on the outer membrane and are involved in adherence to host epithelial cells . Adhesion to host tissues is critical for invasion and pathogenicity of S . Typhimurium [13] . Type 1 fimbriae are one of the best characterized fimbrial adhesins and are encoded by the fim operon . FimH , a lectin-like protein , directly binds to high mannose oligosaccharides conjugated to surface glycoproteins of eukaryotic cells [14 , 15] . Another fimbrial operon , std , encodes the π-class Std fimbriae , which have been described to bind terminal α-1 , 2 fucose residues [16] . The expression of bacterial adhesins possibly involved in binding fucosylated host proteins suggests that these fimbriae may facilitate Salmonella to establish or maintain infection in the highly fucosylated large intestine . Here , we investigated the role of host fucosylation in disease development during Salmonella infection using mice with and without expression of the Fut2 gene ( Fut2+/+ and Fut2-/- ) . Taken together , our results demonstrate that Std-fucose interaction contributes to S . Typhimurium persistence and inflammation .
To test the hypothesis that expression of Fut2 influences host susceptibility to enteric pathogens , a model of S . Typhimurium-induced colitis was utilized . Fut2+/+ and Fut2-/- littermates were treated with streptomycin , and 24 hours later , infected with wild-type S . Typhimurium . One day post infection ( p . i . ) , the cecal tissue of Fut2-/- mice contained more S . Typhimurium than Fut2+/+ mice ( S1A Fig ) in agreement with the observations of Goto and colleagues [17] . However , in contrast to their results , we found the total cecum weight and histopathology scores ( S1B–S1D Fig ) were comparable between Fut2+/+ and Fut2-/- mice . Under most conditions wild-type Salmonella kill C57BL/6 mice within approximately one week . Therefore , in order to follow the infection to later time points mice were infected with the S . Typhimurium ΔaroA mutant strain which is attenuated for systemic disease but causes extensive intestinal inflammation [18] . There was no significant difference in bacterial colonization or resulting inflammation of Fut2+/+ and Fut2-/- mice on day 1 p . i . ( S1 Fig ) or on day 3 p . i . ( S2A–S2C Fig ) . However , on day 7 and day 14 p . i . , a significantly reduced Salmonella burden in the intestine of Fut2-/- compared to Fut2+/+ mice was detected ( Fig 1A , S2A–S2C Fig ) . Furthermore , 7 days p . i . , the histopathological changes in the colon were significantly less severe in Fut2-/- mice compared to Fut2+/+ animals ( Fig 1B and 1C ) . Notably , the colons of infected Fut2+/+ mice were characterized by a higher number of detached epithelial cells within the colon lumen , increased inflammatory cell infiltration within the mucosa , and stronger submucosal edema . Additionally , the levels of the inflammation-associated marker lipocalin-2 were quantified in the large intestine after S . Typhimurium infection . The concentration of lipocalin-2 in the colon and cecum 7 days p . i . were significantly higher in Fut2+/+ mice in comparison to Fut2-/- ( Fig 1D , S2D Fig ) . Next , colon tissue sections were analyzed by immunohistochemical staining and subsequent quantification of CD68- and MPO-positive cells , which represent macrophages and neutrophils , respectively . Consistent with elevated histopathological scores , significantly higher numbers of recruited neutrophils and macrophages were detected in the colon tissue of Fut2+/+ mice compared to Fut2-/- mice ( Fig 2A–2C ) . In addition , a significantly stronger infiltration of CD4+ T lymphocytes in the colonic lamina propria of Fut2+/+ mice compared to Fut2-/- mice was detected by immunofluorescence staining and by flow cytometry ( Fig 2A and 2D ) . No statistically significant differences were found in the numbers or infiltrate composition with respect to cytotoxic T lymphocytes , B lymphocytes , or dendritic cells by flow cytometric quantification of CD8+ , CD19+ , and CD11chi cells , respectively ( S3 Fig ) . To summarize , Fut2-expressing mice exhibited higher bacterial load in the intestine at later time points , which was also associated with an increase in inflammation assessed by histopathology and lipocalin-2 levels , demonstrating that Fut2-mediated fucosylation in the intestine plays an important role in Salmonella-triggered inflammation and colonization of the intestine . The Fut2 protein facilitates intestinal epithelial fucosylation by catalyzing the addition of L-fucose residues via an α ( 1 , 2 ) linkage to the terminal β-D-galactose residue of glycans . The std operon of S . Typhimurium encodes a fimbrial adhesin known to be important for the attachment to fucosylated structures on intestinal epithelial cells [16] . To investigate the role of fucose-Std fimbriae interaction during Salmonella adherence , human intestinal epithelial cell lines HT29-MTX-E12 and Caco-2 were utilized . HT29-MTX-E12 are colon epithelial cells that differentiate into goblet-like cells and produce mucus after three weeks of in vitro culture [19] . Ulex europaeus agglutinin I ( UEA-I ) lectin staining revealed extensive fucosylation of cell surface and mucus in the differentiated HT29-MTX-E12 cells in contrast to undifferentiated HT29-MTX-E12 cells . Wheat germ agglutinin ( WGA ) lectin staining for the ubiquitously expressed N-acetylglucosamine was positive in both differentiated and undifferentiated HT29-MTX-E12 cells ( Fig 3A ) . Previous studies have shown that bacterial expression of Std fimbriae is a subject of complex and tight regulation , both in vivo and in vitro [20 , 21] . Only a very small proportion of the Salmonella population express std fimbriae in vitro [22] and the std operon is completely absent in E . coli . Therefore , to analyze the role of Std fimbriae in vitro , an inducible expression plasmid containing the Salmonella stdABCD operon encoding the structural genes of Std fimbriae was transformed into a common laboratory E . coli K-12 and afimbriated E . coli ORN172 . Upon induction with anhydrotetracycline , Std fimbriae were expressed by E . coli ( referred as E . coli StdON ) as confirmed by flow cytometry and Western blotting ( Fig 3C and 3D ) . HT29-MTX-E12 cells were infected with either std-expressing ( E . coli StdON ) or non-expressing bacteria ( E . coli StdOFF ) . In contrast to E . coli StdOFF , only E . coli StdON strain showed increased adherence to differentiated HT29-MTX-E12 cells ( Fig 3B ) . Importantly , expression of Std fimbriae had no effect on adhesion of E . coli StdON to undifferentiated HT29-MTX-E12 cells which do not contain fucosylated glycoproteins ( Fig 3B ) . Similarly , E . coli ORN172 StdON bacteria adhered significantly better to Caco-2 cells compared to E . coli lacking Std expression . Addition of fucose-binding UEA-1 lectin to the cells prior to infection abrogated the adhesion of the std-expressing strain . In contrast , addition of dolichus biflorus aggluttinin ( DBA ) , which binds to N-acetylgalactosamine , did not affect binding of E . coli ORN172 StdON to Caco-2 cells ( S4A and S4B Fig ) . Atomic force microscopy showed Std piliation of E . coli ORN172 StdON and the absence of pili in the empty vector control bacteria ( S4C Fig ) . In conclusion , Std fimbriae are important for binding α ( 1 , 2 ) -fucosylated residues on cell lines corroborating the results by Chessa and colleagues [16] . Next , Std-dependent bacterial adherence to primary epithelial cells was investigated . To this end , primary intestinal epithelial cells were isolated from Fut2+/+ mice and cultivated as three-dimensional organoids in matrigel . These enteroids were expanded and seeded onto transwell filters resulting in the formation of a 2D monolayer consisting of various primary epithelial cell types . Monolayer barrier integrity and the degree of differentiation were evaluated by measuring transepithelial electrical resistance . Polarized monolayers were infected with E . coli StdON and StdOFF bacteria and adherence was analyzed by immunofluorescence . We counted the number of E . coli bacteria attached to UEA-1-positive and -negative cells . E . coli StdON bacteria were primarily associated with fucosylated cells , while E . coli StdOFF adhered equally to fucosylated and non-fucosylated cells ( Fig 4A ) . Furthermore , a significantly higher number of the α ( 1 , 2 ) -fucose-associated E . coli StdON cells compared to the α ( 1 , 2 ) -fucose-associated E . coli StdOFF bacteria was detected ( Fig 4B ) . Overall , this data demonstrate that std-expressing bacteria preferentially bind to fucosylated cells . Using cut sections of the cecum of CBA/J mice , it was previously demonstrated that purified Std fimbriae of S . Typhimurium are able to bind terminal α ( 1 , 2 ) -fucose residues in the mucosa [16] . However , the functional consequences of this interaction for disease development , as well as the extent of Std fimbriae expression in vivo are not known . To assess Std fimbriae production in vivo , stdA gene expression was first examined using RT-qPCR . Similar levels of stdA gene expression were detected in the colon of both Fut2+/+ and Fut2-/- mice infected with S . Typhimurium ΔaroA strain ( S5A Fig ) . In order to determine whether the absence of std or the presence of intestinal fucosylated glycans affects expression of fucose or 1 , 2-propanediol utilization pathways we quantified levels of fucI and pduBC by RT-qPCR . We saw comparable levels of these genes expressed in either mouse strain in std-containing and std-deficient bacteria ( S5B and S5C Fig ) . In order to look more closely at the spatial regulation of Std expression , we used reporter strains of S . Typhimurium containing a stdAstop::gfp fusion [22] and staining of tissue sections with anti-Std serum . Std was observed to be specifically expressed in the lumen of the large intestine of both Fut2+/+ and Fut2-/- mice on day 1 ( S6 Fig ) and day 7 p . i . ( Fig 5 and S7 Fig ) . In contrast , Std-expressing Salmonella were not observed after invasion of the mucosa ( Fig 5 , S6 and S7 Figs ) . These data demonstrate that there is a tight spatial regulation of Std-expression whereby Std fimbriae are expressed prior to invasion of the large intestine . Fut2-/- mice lack terminal fucose on intestinal epithelium [23] . To test whether epithelial fucosylation directly affects Salmonella colonization via interaction with Std fimbriae , the stdA mutation was transferred into the S . Typhimurium ΔaroA background strain . This mutant strain lacked functional Std fimbriae ( S . Typhimurium ΔaroAΔstdA ) yet had the same growth rate and motility as the parental S . Typhimurium ΔaroA strain ( S8 Fig ) . Fut2+/+ and Fut2-/- mice were then infected with S . Typhimurium ΔaroAΔstdA via oral gavage for 7 days . In contrast to S . Typhimurium ΔaroA ( Fig 1 ) , the S . Typhimurium ΔaroA ΔstdA strain colonized the colon of Fut2+/+ and Fut2-/- mice to similar levels ( Fig 6A ) . In addition , histopathology scores showed similar intestinal inflammation of Fut2+/+ and Fut2-/- mice . H&E staining of colonic tissue showed moderate numbers of necrotic epithelial cells , mild inflammatory cell infiltration within intestinal mucosa and mild submucosal edema ( Fig 6B and 6C ) , which was similar in both mouse genotypes . In addition , lipocalin-2 concentrations were similar in colons of both Fut2+/+ and Fut2-/- mice after infection with S . Typhimurium ΔaroA ΔstdA ( Fig 6D ) . In order to further explore whether Std fimbriae play a role in the colonization of S . Typhimurium in the presence or absence of host intestinal fucosylation , competitive index ( CI ) experiments were performed by orogastrically infecting both Fut2+/+ and Fut2-/- mice with equal numbers of S . Typhimurium ΔaroA and S . Typhimurium ΔaroA ΔstdA . Fecal pellets were collected at 1 , 3 , and 5 days p . i . and bacterial counts of both strains were determined . After 7 days , the bacterial load in intestinal tissues and luminal content was determined and the CI ratio of the two strains was calculated . Interestingly , in cecum and colon , S . Typhimurium ΔaroA significantly outcompeted the isogenic Std-deficient strain in Fut2+/+ mice , but not in Fut2-/- mice ( Fig 7 ) . Accordingly , CI results from feces at day 1 , 3 , and 5 showed that S . Typhimurium ΔaroA outcompeted the isogenic ΔstdA mutant in Fut2+/+ mice ( S9 Fig ) . Taken together , our in vivo data demonstrate that Std fimbriae are important for Salmonella colonization , persistence , and induction of inflammation in a fucosylated host environment .
Variation in human glycosylation influences various metabolic diseases , cancers , inflammatory diseases , and susceptibility to infectious pathogens . Genome-wide association studies show that FUT2 nonsense polymorphisms are associated with increased risk for Crohn’s disease [10] and primary sclerosing cholangitis [24] . Genetic variation in FUT2 is also linked to susceptibility to infections with bacterial and viral pathogens including Helicobacter pylori [25] , norovirus [26 , 27] , Enterotoxigenic E . coli [5] , and progression of HIV [28] . In this study , we investigated the role of Fut2 expression for S . Typhimurium infection and found that Std fimbriae-fucose interaction was important for Salmonella-induced inflammation and colonization . Host mucosal glycans can influence the susceptibility to infection indirectly or directly . Indirectly , glycan-dependent differences in microbiota composition may contribute to the susceptibility to infection with a particular pathogen . For example , we previously reported an influence of the histo-blood group related glycosyltransferase gene B4galnt2 on host susceptibility to S . Typhimurium infection . The expression of B4galnt2 in the gut results in differences in microbial composition which in turn affect the extent of Salmonella colonization , and hence , disease pathology [29] . Many complex carbohydrates degraded by the intestinal microbiota produce metabolites that can be utilized by Salmonella Typhimurium and Clostridium difficile thereby facilitating their expansion within the gut [4] . Host glycans can also directly influence a host’s susceptibility to infection by modulating bacterial attachment to host tissues . Many bacteria produce specific adhesins which bind to host glycans . For example , the H . pylori adhesin BabA mediates adherence to the gastric mucosa of individuals with fucosylated ABO ( H ) /Lewis b blood group antigens [30 , 31] . Norovirus ( strain GII . 4 ) and rotavirus ( strains with spike protein VP8 ) encode adhesins which bind α ( 1 , 2 ) -fucosylated glycans resulting in increased susceptibility in individuals with a secretor phenotype [32 , 33] . While Fut2 expression in the small intestine is inducible by ILC3-derived IL-22 [17] , the large intestine is constitutively fucosylated [8] . Mice lacking Fut2 are more susceptible to S . Typhimurium infection at an early time point post infection , as demonstrated by Goto and colleagues [17] and confirmed in the present study . In contrast , we demonstrate that at later time points , a lack of Fut2 expression is associated with decreased intestinal colonization , pathology , and inflammatory responses . It has been hypothesized that at one day post S . Typhimurium infection , α ( 1 , 2 ) -fucose-containing glycans secreted from goblet cells may interfere with the attachment of Salmonella to intestinal epithelial cells , although this has not been proven so far [34] . Our study suggests that at later time points post infection , Salmonella exploits α ( 1 , 2 ) -fucose-containing glycans present in the intestine of Fut2+/+ mice to their advantage . Fucose and its metabolic products such as 1 , 2-propanediol can be utilized by S . Typhimurium as carbon and energy sources [35] . In vivo , during intestinal inflammation , 1 , 2-propanediol is generated and serves as a nutrient source for Salmonella [36] . Our data do not show any differences in the expression of fucI and pduBC in Fut2+/+ and Fut2-/- mice , even in the absence of stdABCD genes , suggesting that these pathways are not regulated by Std fimbriae . It was previously shown that purified Std fimbriae of S . Typhimurium can bind terminal α ( 1 , 2 ) -fucose residues [16] . Salmonella enterica encodes up to 13 fimbrial and at least 7 non-fimbrial adhesins depending on the serovar [13 , 37] . The production of adhesins involves complex and tight regulation since inappropriate expression could be detrimental for bacterial colonization and pathogenesis . While the majority of adhesins are expressed in temporal and spatially highly controlled manner during animal infections , they are often not produced under laboratory conditions [20 , 38] . In vitro expression of the S . Typhimurium std operon is bistable resulting in the emergence of a minor subpopulation of Std-positive cells [22] . Therefore , to investigate the role of Std fimbriae in vitro we took advantage of an Std-expressing E . coli strain . Our in vitro infection experiments revealed enhanced attachment of Std-expressing E . coli to fucosylated cell lines , which could be specifically blocked by the addition of the α ( 1 , 2 ) -fucose-binding lectin UEA-I . These observations confirm the findings of Chessa and colleagues [16] and were further corroborated by preferential binding of Std-expressing E . coli to fucosylated cells in primary epithelial monolayers . Altogether , bacteria expressing Std fimbriae exhibited increased adhesion to human cell lines and murine intestinal crypt organoids when terminal α ( 1 , 2 ) -fucose was present . The extent and localization of Std fimbriae expression in vivo was not previously known . Notably , we observed the spatial expression pattern of Std fimbriae in vivo . In this work , we present , for the first time , evidence that Std fimbriae are differentially produced during intestinal infection . Std-positive Salmonella were found predominantly in the intestinal lumen , which corroborates the existence of a regulatory fimbrial switch . Previous work has shown that Std fimbriae are required for the long-term S . Typhimurium colonization of CBA mice [39] . In agreement with these data , we show that Std fimbriae are important for persistence in C57Bl/6 mice and demonstrate that this is strictly dependent on the ability of Std fimbriae to adhere to fucosylated host glycoproteins or glycolipids in the large intestine . While the expression of α ( 1 , 2 ) -fucose on enterocytes and goblet cells is facilitated by Fut2 , M cells in the follicle associated epithelium overlying Peyer’s patches [40] and Paneth cells [41] are α ( 1 , 2 ) -fucosylated via Fut1 . Thus , based on our CI data , we can conclude that Fut2-dependent fucosylation is important for Salmonella persistence only in the large intestine . Furthermore , we present evidence that Std fimbriae are also important for the induction of host intestinal pathology and inflammation in a Fut2-dependent manner . In conclusion , our results demonstrate a substantial role for glycosylation of the intestinal mucosa in the susceptibility to S . Typhimurium infection . Std fimbriae binding of terminal α-1 , 2 fucose residues mediate bacterial adherence to host glycoproteins or glycolipids . Taken together , our results conclusively demonstrate that host fucosylation in the intestine is exploited by S . Typhimurium during the course of infection in a mechanism that requires Salmonella expression of Std fimbriae .
B6 . 129X1-Fut2tm1Sdo/J ( Fut2-/- ) [42] mice were purchased from the Jackson Laboratory and intercrossed with wild-type C57BL/6J ( Fut2+/+ ) mice . Mice were backcrossed for 14 generations . Heterozygous breeding pairs produced litters of mixed genotypes . Slc11a1 ( Nramp1 ) is an important resistance gene for S . Typhimurium . 129X1 mice harbor Slc11a1 resistant alleles while in C57Bl/6J mice a point mutation results in Slc11a1 sensitive alleles . The Slc11a1 genotype of mice was verified as described [[43]] using a common reverse primer and forward primers ( S2 Table ) specific for the sensitive and resistant allele , respectively . All mice were homozygous for the Slc11a1 sensitive allele . Mice were housed together under specific pathogen-free conditions in individually ventilated cages ( IVC ) . Standard chow and water were provided ad libitum . Experiments were conducted in the animal facilities of the University of Kiel and Hannover Medical School in Germany . Animal experiments were conducted in direct accordance with the German Animal Protection Law consistent with the ethical requirements and approval of the Animal Care Committee of the Ministry of Energy , Agriculture , the Environment and Rural Areas of Schleswig-Holstein , Germany ( protocol # V244-7224 . 121 . 3 ( 99-10/10 ) ) and by the Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit ( protocol # 33 . 12-42502-04-16/2071 ) . S . Typhimurium SL1344 ( S . Tm ) [44] , S . Tm ΔaroA [18] were grown at 37°C with shaking in lysogeny broth ( LB ) supplemented with 100 μg/ml streptomycin . S . Tm ΔaroAΔstdAB ( referred to as S . Tm ΔaroAΔstdA ) double mutant was generated by P22 phage transduction of the ΔstdAB deletion from the S . Tm NCTC 12023 ΔstdAB [38] to the S . Tm SL1344 ΔaroA background . S . Tm ΔaroA stdAstop::gfp strain was created by P22 transduction of the ΔaroA deletion into the parental S . Tm stdAstop::gfp strain [22] . For in vivo infection , S . Tm ΔaroAΔstdA and S . Tm ΔaroA stdAstop::GFP strains were grown overnight in LB broth supplemented with kanamycin 50 μg/ml at 37°C . The E . coli Std strain containing the anhydrotetracyclin ( AHT ) -inducible Salmonella stdABCD operon was generated by electroporation of the plasmid p4394 [45] into E . coli Turbo ( New England Biolabs ) K-12 strain ( referred as WT ) . For in vitro infection experiments , E . coli stdABCD was grown overnight in LB broth supplemented with carbenicillin 50 μg/ml at 37°C and then grown to logarithmic phase in the presence ( StdON ) or absence ( StdOFF ) of AHT ( IBA , 100 ng/ml ) . The non-fimbriated E . coli ORN172 strain [38] harboring the empty vector or a plasmid for expression of the Salmonella stdABCD operon under control of the tetR PtetA expression cassette were grown in LB broth . Expression was induced by addition of 100 ng/ml anhydrotetracycline ( AHT ) . Imaging of bacteria by atomic force microscopy ( AFM ) was performed as previously described [45] . Fut2-/- and wild-type ( Fut2+/+ ) littermates were pretreated by oral gavage with 20 mg of streptomycin ( Sigma-Aldrich ) 24 hours before infection . Mice were orally gavaged with either 3x106 S . Tm , S . Tm ΔaroA , S . Tm ΔaroAΔstdA , S . Tm stdAstop::gfp , or S . Tm ΔaroA stdAstop::gfp in 100 μl HEPES buffer ( 100 mM , pH 8 . 0; PAA Laboratories ) . Control mice ( mock infection ) were given 100 μl HEPES buffer . To enumerate luminal and tissue-invaded bacteria , colon and cecum tissues were harvested and the intestinal contents were separated from the tissues . Tissues were then treated with PBS containing 100 μg/ml gentamicin at 4°C for 30 minutes to kill bacteria on the tissue surface . Intestinal tissues and intestinal contents were homogenized , serially diluted , and plated on LB agar containing streptomycin ( 100 μg/ml ) . Fut2+/+ and Fut2-/- mice were pre-treated with 20 mg of streptomycin and infected with a mixture of S . Tm ΔaroA and S . Tm ΔaroAΔstdA strains ( 1:1 , 3x106 total bacteria per mice ) . During the infection , fecal pellets were collected at days 1 , 3 , and 5 p . i . At 7 days p . i . , mice were sacrificed and bacterial loads in cecum tissue , cecum content , colon tissue , colon content , and ileum were enumerated . Feces and intestinal tissues were homogenized and plated on selective LB agar plates supplemented with either streptomycin ( 100 μg/ml ) alone ( to determine total Salmonella load ) or with streptomycin ( 100 μg/ml ) and kanamycin ( 50 μg/ml ) ( to enumerate S . Tm ΔaroAΔstdA ) . The number of S . Tm ΔaroA was calculated by subtracting the CFU counts of S . Tm ΔaroAΔstdA from the total Salmonella counts . Competitive index ( CI ) was calculated as the ratio of ( ΔaroA / ΔaroAΔstdA ) at the time of sampling divided by ( ΔaroA / ΔaroAΔstdA ) of the inoculum . Motility was assessed by inoculating motility agar plates ( 10 g/l tryptone , 5 g/l NaCl , and 0 . 3% Bacto-agar ) with saturated bacterial cultures grown overnight in LB broth . Motility halos were compared after incubation at 37°C for 6 hours . Organs were fixed in 10% formalin , dehydrated with ethanol , and embedded in paraffin . Paraffin sections were stained with hematoxylin-eosin ( H&E ) according to standard laboratory procedures . Histological scores of ceca and colons were determined as previously described [46] . Briefly , pathological changes were assessed by evaluating the presence of necrotic epithelial cells and neutrophils in lumen; desquamation and ulceration in surface epithelial cells; crypt abscesses; infiltrating inflammatory and immune cells in mucosa and submucosa layer; and the formation of edema in submucosa layer . Human colon epithelial clonal cell line , HT29-MTX-E12 [47 , 48] ( a kind gift from Marguerite Clyne , University College Dublin ) , and the colorectal carcinoma cell line , Caco-2BBe1 ( ATCC CRL-2102 ) , were grown in DMEM supplemented with 10% fetal bovine serum ( Biochrom ) and 1% MEM non-essential amino acids solution ( Gibco , Life Technologies ) . Cells were seeded in 24 well plates and incubated at 37°C in a humidified 5% CO2 atmosphere . Cells were grown for 7 days ( Caco-2-BBe1 ) or 21 days ( HT29-MTX-E12 ) to achieve differentiation of the monolayers . Cells were then infected with E . coli StdOFF or E . coli StdON , E . coli ORN172 StdOFF or E . coli ORN172 StdON for 30 min at an MOI of 50 . For quantification of adherence , cells were washed four times with PBS and lysed in PBS containing 1% ( v/v ) Triton X-100 . The number of adherent bacteria was determined by serial dilutions plating . Where indicated , 30 min prior to infection , cells were incubated with medium in presence of 0 . 3 mM UEA-I or DBA lectins ( CosmoBio ) at 37°C in a humidified 5% CO2 atmosphere . Primary colonic and ileal crypts were isolated from Fut2+/+ mice as described [49] with modifications . Briefly , mice were sacrificed by cervical dislocation . Intestines were opened longitudinally , cut into small pieces and washed three times with 5 ml of ice-cold DPBS . Tissues were then incubated in 10 ml of ice-cold crypt chelating buffer ( 10 mM EDTA in DPBS ) for 90 min on an orbital shaker . The supernatant was discarded and the settled tissue fragments were resuspended twice in 5 ml ice-cold DPBS . Crypts were centrifuged for 5 min at 800 rpm at 4°C and pellets were resuspended in 1 ml ice-cold DPBS . About 100 crypts were resuspended in 25 μl organoid medium ( Advanced DMEM/F12 medium ( Thermo Fischer Scientific ) supplemented with 2 mM GlutaMax , 50% L-WRN-Supernatant ( ATCC CRL3276 ) , 10 mM HEPES , 100 U/ml penicillin , 100 μg/ml streptomycin , B27 supplement , 50 ng/ml recombinant mouse epidermal growth factor ( rm EGF ) , 500 nM A83-01 ( Tocris ) , 10 μM SB202190 ( Tocris ) , 10 nM Gastrin I ( Tocris ) , 1 mM N-Acetyl-L-cysteine ( Sigma ) , and 10 μM Y27623 ( Tocris ) ) . 25μl Matrigel ( Corning ) was added into a well of a pre-warmed 24-well plate . The plate was incubated for 0 . 5 h in a 37°C incubator with 5% CO2 to allow complete polymerization of the Matrigel . Crypts were covered with 500 μl of the organoid medium . To form 2D monolayers , 3D organoids were resuspended in ice-cold PBS and centrifuged at 1500 rpm for 10 min at 4°C . Pellets were resuspended in 1 ml warm 0 . 05% trypsin/EDTA and incubated for 5 min at 37°C in a water bath . Organoids were dissociated by pipetting and washed with ice-cold DMEM/10% FCS and resuspended in monolayer medium ( Advanced DMEM/F-12 , 50% L-WRN-Supernatant , 20% fetal bovine serum , 2 mM L-glutamine , 100 U/ml penicillin , 0 . 1 mg/ml streptomycin , 10 μM Y-27632 , 50 ng/ml rm-EGF ) . Cell suspensions were seeded onto Transwell permeable supports ( polyester; 6 . 5 mm diameter; 0 . 4 μm pore size; Corning ) that had been coated for 2 h at 37°C with Matrigel ( diluted 1:40 in PBS ) . Monolayer medium was replaced every two days and monolayer barrier integrity was evaluated by measuring transepithelial electrical resistance ( TEER ) using a volt-ohmmeter ( Millipore ) . On day 5 after seeding , medium was changed to differentiation medium ( Advanced DMEM/F-12 , 5% L-WRN-Supernatant , 20% fetal bovine serum , 2 mM L-glutamine , 50 ng/ml rm-EGF , 5 μM DAPT ) . For the next two days , differentiation medium was changed every day and TEER was measured . 2D monolayers were infected with either E . coli StdON or E . coli StdOFF ( 7 x 107 bacteria per Transwell ) , incubated for 30 min at 37°C , washed four times with PBS , and fixed with 4% paraformaldehyde ( PFA ) . Adherent fucose-associated and non-associated bacteria were counted microscopically , using at least 20 fields of view ( FOV ) per sample . Formalin-fixed paraffin-embedded tissue sections ( 5 μm ) were deparaffinized and rehydrated . Heat-induced epitope retrieval was performed using 10 mM sodium citrate buffer ( pH 6 . 0 ) and blocking was achieved using 2% normal goat serum ( NGS ) . The following antibodies were used for immunohistochemistry ( see S1 Table for a full description ) : anti-StdA serum [38] , Salmonella O Antiserum Group B ( BD Difco ) , anti-GFP ( DSHB ) , CD3 ( Abcam ) , CD68 ( Abcam ) , myeloperoxidase ( MPO ) ( Thermo Fisher Scientific ) , and fluorescently-labeled secondary antibodies ( Invitrogen ) . Counterstaining of nuclei was done with 4 , 6-Diamidin-2-phenylindol ( DAPI ) ( Invitrogen ) . HT29-MTX-E12 and Caco-2 Bbe1 cells were seeded on coverslips in 24 well plates and fixed with 4% PFA before and after differentiation . Blocking of non-specific binding was done using 2% NGS . Fluorescently-labeled lectins UEA-1 ( Ulex Europaeus agglutinin-1 ) ( CosmoBio ) and WGA ( wheat germ agglutinin ) ( Vector laboratories ) were used to visualize α ( 1 , 2 ) -fucosylation and the presence of sialic acid / N-acetylglucosaminyl residues , respectively . Fixed primary epithelial cell monolayers were stained with the anti-E . coli antibody ( Abcam ) , UEA-1 lectin ( CosmoBio ) , and DAPI . Images were obtained on a Zeiss Apotome . 2 microscope using AxioVision 4 . 9 . 1 software ( Zeiss ) and on a Leica DMi8 confocal laser scanning microscope using LAS X 3 . 3 . 0 . 16799 software ( Leica ) . Brightness and contrast were adjusted using ImageJ 1 . 52e software . Isolation of colonic lamina propria cells was achieved using the Lamina Propria Dissociation Kit ( Miltenyi Biotec ) according to the manufacturer’s protocol . Leukocyte isolation was performed with 40% / 80% discontinuous Percoll gradient ( GE Healthcare ) . Cells were incubated with FcγR blocking reagent ( rat anti-mouse CD16/CD32 , BD Biosciences ) for 30 minutes on ice prior to incubation with the other antibodies . Antibodies used for flow cytometry analysis are listed in S1 Table . To detect expression of Std fimbriae , flow cytometry was performed as previously described [16] with modifications . In brief , approximately 5x108 bacteria were fixed with 10% formalin and incubated at room temperature for 20 minutes . After washing with PBS , cells were resuspended in 2% NGS diluted in PBS and incubated at room temperature for 30 minutes . Polyclonal rabbit anti-StdA serum [38] was added to the cell suspensions following incubation at room temperature for 30 min . After washing with PBS , fluorescently labeled secondary antibodies ( Invitrogen ) were added . Flow cytometry was performed using a MACSQuant Analyzer 10 ( Miltenyi Biotec ) . The data were analyzed using FlowJo v . 10 software ( TreeStar ) . Supernatants from the organ homogenates were collected and stored at −20°C . Lipocalin-2 levels were detected using mouse lipocalin-2/NGAL DuoSet ELISA ( R&D Systems ) according to the manufacturer’s protocol . Absorbance was measured using a Synergy HTX microplate reader and acquired using Gen5 software ( Biotek ) . E . coli StdOFF and E . coli StdON were grown in LB broth supplemented with carbenicillin ( 50 μg/ml ) and to induce std fimbrial expression anhydrotetracycline ( 100 ng/ml ) at 37°C until an OD600 of 0 . 6 was reached . 108 bacteria were pelleted , resuspended in PBS , mixed with an equal volume of Laemmli buffer supplemented with 10% DTT , and boiled for 10 min . These whole-cell lysates were spun down and supernatants were loaded immediately onto a SDS-PAGE gel ( 15% ) . Proteins were transferred to Hybond-P 0 . 45 PVDF ( Amersham ) membranes using a Trans-Blot semi-dry transfer cell ( Bio-Rad ) . After blocking with Roti-block ( Carl Roth ) , membranes were incubated first with anti-StdA [38] serum ( Humphries et al . , 2003 ) diluted 1:500 in blocking buffer , and then with a goat anti-rabbit-HRP conjugate , and finally with Pierce ECL Western Blotting Substrate ( Thermo Fisher Scientific ) . Images were obtained using the ImageQuant LAS 4000 system ( GE Healthcare ) . Feces of the infected Fut2+/+ and Fut2-/- mice ( day 7 p . i . ) was immediately stored in RNAlater ( Ambion ) . Total RNA was extracted using the High Pure RNA Tissue Kit ( Roche ) and reverse transcription was performed with the cDNA Synthesis Kit ( Roche ) in accordance with the manufacturer’s instructions . Quantitative real-time PCR ( qPCR ) was performed on a CFX96 Real-Time PCR Detection System ( Bio-Rad ) using the Power SYBR Green PCR Master Mix ( Applied Biosystems ) . Gene-specific primers are listed in S2 Table . Data were normalized to the house-keeping gene rpoD and analyzed by ΔΔCt method [50] with median values of Fut2+/+ mice infected with S . Tm ΔaroA as calibrators . All data were analyzed using GraphPad Prism V7 . 0d software . Statistical analyses were performed using one-way analysis of variance followed by Tukey’s multiple comparison test or Wilcoxon-Mann-Whitney test as indicated . Graphs display the mean values ± SD , unless stated otherwise . Competitive index data were analyzed by the Wilcoxon signed-rank test by comparing medians with a hypothetical value of 1 . | The intestinal epithelium is a crucial biological interface , interacting with both commensal and pathogenic microorganisms . It’s lined with heavily glycosylated proteins and glycolipids which can act as both attachment sites and energy sources for intestinal bacteria . Fut2 , the enzyme governing epithelial α1 , 2-fucosylation , has been implicated in the interaction between microbes and intestinal epithelial cells . Salmonella is one of the most important bacterial gastrointestinal pathogens affecting millions of people worldwide . Salmonella possesses fimbrial and non-fimbrial adhesins which can be used to adhere to host cells . Here we show that Salmonella expresses Std fimbriae in the gastrointestinal tract in vivo and exploit Std fimbriae to bind fucosylated structures in the mucus and on the intestinal epithelium . Furthermore , we demonstrate that the Std fimbriae-fucose interaction is necessary for bacterial colonization of the intestine and for triggering intestinal inflammation . These data lend new insights into bacterial adhesion-epithelial interactions which are essential for bacterial pathogenesis and key factors in determining tissue tropism and host susceptibility to infectious disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"pili",
"and",
"fimbriae",
"salmonellosis",
"epithelial",
"cells",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"enterobacteriaceae",
... | 2019 | Std fimbriae-fucose interaction increases Salmonella-induced intestinal inflammation and prolongs colonization |
A spatially-explicit , stochastic model is developed for Bahia bark scaling , a threat to citrus production in north-eastern Brazil , and is used to assess epidemiological principles underlying the cost-effectiveness of disease control strategies . The model is fitted via Markov chain Monte Carlo with data augmentation to snapshots of disease spread derived from a previously-reported multi-year experiment . Goodness-of-fit tests strongly supported the fit of the model , even though the detailed etiology of the disease is unknown and was not explicitly included in the model . Key epidemiological parameters including the infection rate , incubation period and scale of dispersal are estimated from the spread data . This allows us to scale-up the experimental results to predict the effect of the level of initial inoculum on disease progression in a typically-sized citrus grove . The efficacies of two cultural control measures are assessed: altering the spacing of host plants , and roguing symptomatic trees . Reducing planting density can slow disease spread significantly if the distance between hosts is sufficiently large . However , low density groves have fewer plants per hectare . The optimum density of productive plants is therefore recovered at an intermediate host spacing . Roguing , even when detection of symptomatic plants is imperfect , can lead to very effective control . However , scouting for disease symptoms incurs a cost . We use the model to balance the cost of scouting against the number of plants lost to disease , and show how to determine a roguing schedule that optimises profit . The trade-offs underlying the two optima we identify—the optimal host spacing and the optimal roguing schedule—are applicable to many pathosystems . Our work demonstrates how a carefully parameterised mathematical model can be used to find these optima . It also illustrates how mathematical models can be used in even this most challenging of situations in which the underlying epidemiology is ill-understood .
Mathematical models of plant disease can be used to screen and assess control strategies [1]–[10] . Although work on plants is not subject to the ethical concerns that hamper experimentation targeting pathogens of animal or human hosts , mathematical modelling nevertheless becomes particularly compelling for plant diseases when logistic constraints mean that experimentation would be costly or difficult . This situation is exemplified by diseases caused by pathogens with epidemiology necessitating long experiments to yield useful data [11]–[13] , pathogens causing symptoms that are difficult to detect [14] , [15] , pathogens with epidemiology that is ill-understood [16] , [17] , and/or pathogens that would require experimental trials in the vicinity of susceptible commercial growing operations [18] , [19] . Here we develop a model of Bahia bark scaling of citrus ( BBSC ) on grapefruit , a pathosystem subject to each of these logistical constraints . BSSC has been endemic to north-eastern Brazil since the 1960s [20] , but its etiology remains unknown [21] . We use Markov chain Monte Carlo with data augmentation [22] to fit a spatially-explicit , stochastic , epidemiological model to a data-set charting the spread of BBSC through a small experimental grove . We go on to alter the host topology and parameters in this model to use it to assess the efficiency and cost-effectiveness of control at the scale of a typical grove as used in citrus production in Brazil . As little is known of the putative BBSC pathogen , and even less about any potential vector , it is difficult to reliably estimate the efficacy of any chemical [23] or biological [24] control . We therefore concentrate on cultural strategies [25] , and focus on the effectiveness of reducing the density of planting [26] and of roguing [10] ( i . e . searching for and removing infected plants ) . The spread of plant pathogens is typically localised , and so it is intuitive that the progression of disease through a host population will be affected by planting density . Direct as well as indirect effects of host density on disease incidence have been proposed [26] , and lower host densities are almost always associated with lower levels of disease [27] . Indeed the “dilution effect” caused by increased distances between pairs of hosts has been suggested to underlie the success of crop mixtures [28] and intercropping [29] , although other more complex mechanisms are thought to be involved in both cases [30]–[33] . However , there are very few models specfically targeting the effect of host density on disease spread . Despite work concentrating on how percolation thresholds can be related to the distance between pairs of nearest neighbours [34] , [35] , tests of that theory have largely been restricted to small-scale model systems [36] , and application to real pathosystems remains in its infancy [37] , [38] . Percolation is also only strictly relevant to systems where spread is restricted to nearest neighbour transmission , although this does map well to the soil-borne pathogens that are the focus of that work . Other work has concentrated on how host density affects invasion thresholds [39] , [40] , but does not provide a clear prescription for how to optimise host densities when disease is able to invade . While there have also been studies showing how the landscape-scale dynamics of disease are conditioned on the configuration and availability of patches of suitable habitat [41] , or fields planted with susceptible varieties [42] , that work offers little at scales relevant to individual farmers or growers . Roguing is commonly used for systemic diseases of high-value or perennial crops [43] , particularly when labour is cheap compared with the cost of chemicals [44] , or for pathogens which cannot be effectively controlled by chemical means [18] , [19] . Viral pathogens for which roguing is practised include cassava mosaic [45] , bunchy top of banana [46] , cocao swollen shoot [47] , [48] , citrus tristeza [49] , plum pox [50] and sweet potato chlorotic stunt [51] , although roguing is also used for bacterial pathogens ( e . g . almond leaf scorch , caused by Xylella fastidiosa [52] ) , and for fungal diseases ( e . g . lettuce drop , caused by Sclerotinia minor [53] ) . The only constraint is that pathogens must cause symptoms that can be detected , either by visual inspection or by diagnostic testing . Roguing has been included in non-spatial mathematical models for a number of years [1] , [2] , [54] , [55] , and more recent work has embedded control by roguing in spatial models of pathogen spread [10] , [56]–[58] , although realistic parameterisation of pathogen dispersal is less common [6]–[8] . Typically these later models have also considered culling , in which all hosts within a particular distance of a symptomatic focal host are removed at the time of control . Some of these models [57] , [58] have explicitly included economics , although the focus has been the cost of treatment ( i . e . the cost of removal of diseased host plants ) . For perennial hosts that are not replanted , however , the cost of detection may , in fact , be more important , since an individual host can be removed at most once , but may be examined for symptoms any number of times . The only model to include detection costs used optimal control theory to show rigorously how to balance the costs of detection and control within a fixed budget [59] , but the mathematical complexity of this procedure necessarily restricted attention to a non-spatial , deterministic model . There are no examples of a model-based approach that optimises the economic aspects of roguing including the cost of detection via a model parameterised to spread data . We have taken advantage of the availability of experimental data for model fitting to frame our analyses specifically in terms of the dynamics and control of BBSC . However , the controls we examine are widely used , and the techniques we use in our modelling and fitting are applicable to a large number of pathosystems . We therefore prefer to think of the BBSC system as a data-driven case study that provides an opportunity to address the following more general questions .
BBSC affects most citrus species and varieties , but is especially severe on grapefruits [60] . Symptoms appear similar to Citrus Psorosis A , and include darkening and thickening of the bark leading to scaling lesions on the trunk and branches , dieback of young branches , and significant gum extrusion . However leaf symptoms on inoculated indicator plants , together with histopathological and molecular studies , indicate BBSC is a distinct disease . The study of Laranjeira et al . [20] resulted in the only published data focusing on BBSC spread ( see Text S1 ) . It demonstrated that the disease is polyetic and naturally transmitted . The speed of disease spread and the pattern of dispersal appear consistent with an insect vector of limited dispersion ability . However the identity of this putative vector is unclear , as is the identity of the pathogen itself [21] . BBSC currently remains restricted to two states in the Brazilian north east , Bahia and Sergipe [21] , [61] . Since dispersal is thought to be localised , the principal risk of an epidemic arising elsewhere in Brazil is likely to occur by transplanting infectious plants . Introduction of BBSC by inadvertent transplantation is certainly possible: Santos et al . [62] have described BBSC symptoms in plants used for budwood in Bahia , Brazil . There is therefore a need to understand whether and how a spatially-isolated epidemic could be effectively controlled . This must be done even though our biological understanding of the epidemiology of BBSC remains limited . We use a spatially-explicit , stochastic , compartmental SEIR model [4] to represent BBSC dynamics at the scale of a grapefruit grove . Individual host plants are categorised by disease status: ( S ) usceptible hosts are uninfected; ( E ) xposed hosts are latently infected , and so are neither symptomatic nor infectious; ( I ) nfected hosts are both infectious and symptomatic; and ( R ) emoved hosts have been removed by control ( Figure 1 ( a ) ) . The E to I transition occurs at rate , corresponding to average latent period ( see also Table S1 ) . Since infectious hosts are always symptomatic in our model , the average incubation period is also . Infected hosts do not appear to suffer increased mortality due to BBSC infection [20] , and so in the absence of control the rate of transition from the I to R compartment is fixed at zero . However , if control by roguing is included in the model , the removal rate is set by how frequently and efficiently infected plants are detected and removed , with rounds of detection and removal according to a schedule that is fixed in advance . Since we work over a twenty year timescale , similar to the typical productive lifespan of an individual citrus host [63] , [64] , we do not attempt to model natural death . We also do not consider replanting of any plants removed by roguing , since this is not common in the Brazilian citrus industry , perhaps due to growers' perception that replanting removed hosts would lead to a heterogeneous grove that would be more difficult to cultivate [63] . The rate of infection of susceptible hosts depends on the disease status of all other hosts in the system . In particular , if host is susceptible at time , then it becomes latently infected ( i . e . transitions to the E compartment ) at rate , where ( 1 ) The summation runs over the set of all ( I ) nfectious hosts , , and denotes the distance from infectious host to susceptible host . The parameter sets the rate of infection . Spatial dependency in spread is controlled by the dispersal kernel , . Here , noting the constant velocity of the epidemic front in the experimental grove [65] , and following exploratory analyses that strongly supported the choice , we used the exponential kernel , normalized in two dimensions ( 2 ) where is the area of susceptible tissue presented by an individual host . The factor of is included since , strictly-speaking , the underlying normalised dispersal kernel is a probability density function , with dimensions of inverse area , meaning the observed rate of infection must be calculated by integration over the area of the recipient plant . Assuming the kernel is constant over this area reduces the integration to a simple multiplication , and so leads to Equation ( 2 ) above [66] , [67] . Since the infection rate then depends entirely on the product in Equation ( 1 ) , we rescale the area of a single host into the infection rate , setting ( 3 ) ( 4 ) ( 5 ) Our model fitting then estimates the value of directly , since it is this product which sets the observed rate of spread of disease in our model . The mean distance of dispersal is [68] . Since we model a grove that initially contains immature plants , and guided by the temporal pattern of disease spread in the experimental grove , we include a delay , , to allow young plants to reach epidemiological maturity [20] , [21] . This delay prevents the disease from spreading for the first units of time , but otherwise does not affect the dynamics of infection in the model . Including this delay is therefore equivalent to considering two age classes of tree in the model: juveniles of age less than , that cannot become infected or transmit infection , and adult trees of age greater than , that are epidemiologically competent . The inclusion of this extra parameter was strongly supported by our model fitting ( see Results and Text S2 ) . Data from the experiment of Laranjeira et al . [20] were used to fit the model . These data consist of successive snapshots over time , tracking the disease status of each host in a small experimental grove . This grove contained 240 grapefruit ( Citrus paradisi Macf . ) plants in 16 rows of 15 . Immature plants were planted at regular 2 m2 m spacing at the start of the experiment , at a closest distance of 5 m from twenty-five BBSC symptomatic adult grapefruit plants arranged in a rectangular lattice at separation 6 m4 m ( see Figure 2a ) . Disease progress was assessed by detailed visual inspection at three monthly intervals for the first five years of the experiment , followed by additional more irregular surveys for two years thereafter . The data consist of the visible disease status of each grapefruit plant in the experimental grove at each survey time; i . e . a series of maps showing which hosts were susceptible and which were ( visibly ) infected on each survey . However , since surveys were separated by at least three months , and because the transition is not visible , exact transition times of individual plants are unknown . We therefore fitted the model in Equations 4 and 5 using Markov chain Monte Carlo with data augmentation to estimate the model parameters of interest ( i . e . and ) [22] , [69] , treating the unobserved times as additional nuisance parameters to be estimated . Posterior distributions for the epidemiological parameters could then be obtained post hoc by marginalization . Further details of the fitting methodology and expressions for likelihood functions are given in the Text S2 . One thousand independent simulations of the model were performed to assess how BBSC would spread in a typical grove ( i . e . 1680 plants at 6 m4 m spacing ) when disease control is not attempted . We ( arbitrarily ) took , and simulated progression over 20 years , a notional productive lifespan of a citrus grove [63] , [64] . Parameter values used in each simulation were drawn randomly from the joint posterior distribution for and as obtained in estimation . The model was simulated using the Gillespie algorithm [70] ( see Text S3 for details ) . The number of plants in the central grove that are susceptible at time is , and the number of plants in the exposed compartment is . We define the number of asymptomatic plants at time as . This corresponds to the number of productive ( i . e . fruit-bearing ) plants at any time . We consider the final number of asymptomatic plants after twenty years , , as a simple composite measure of disease spread , corresponding to the productive trees that remain after accounting for the final size of the epidemic over a 20 year period , and we examined the response of this to values of ranging from 0 . 06% to 2% , i . e . from 1 to 34 initially exposed trees within the central grove . We again used 1000 independent simulations for each initial condition we considered , as we did for each set of parameters in each of the scenarios described below . To test the effect of host density on disease dynamics , the within-row and between-row spacing of trees were altered , while constraining the total number of trees in the central grove to remain fixed at 1680 . The ratio of horizontal to vertical separation was held fixed at throughout . Again we focused on the final number of asymptomatic plants ( ) in a grove with , and considered planting densities from 50 to 500 plants per hectare . While this approach illustrates the effect of inter-host distance on disease spread , it is an oversimplification , since fixing the number of trees at different planting densities corresponds to groves with different areas . To examine the trade-off between disease prevention and productivity we therefore considered the density of asymptomatic trees at years in the central grove as a function of host density , again for . We modelled a programme of scouting for disease symptoms and roguing detected infected plants . This was included in the model by simulating the examination of every surviving plant in the central grove every units of time , and independently detecting symptomatic ( i . e . class I ) plants with probability . Any detected plants were immediately removed . We considered roguing intervals , , between 7 days and 2 years , and took the probability of detection on a round of scouting to be , supported by data from Belasque et al . [71] . Again we assessed the efficacy of control by examining the value of , the number of productive trees in the central grove after twenty years . We considered the responses of to the roguing interval ( ) with fixed , and to with fixed months . We also considered the response of the median value of and of the probability of eradicating the pathogen within twenty years as both and were varied simultaneously . Since the default detection probability is an estimate , we also considered the sensitivity of our results to this choice , by considering the response of the median value of as and were simultaneously varied .
Although the disease initially spreads rather slowly , almost all plants within a typical grove are expected to become symptomatic within 20 years when the initial level of infection ( Figure 1 ( b ) ) . On average 50% of plants become symptomatic within approximately the first 10 years . Spatial snapshots from an arbitrarily chosen run of the model ( Figure 1 ( c ) ) indicate that disease spread is very localised , with infection apparently being transmitted largely ( but not exclusively ) between neighbouring pairs of plants . It also appears to be rather difficult for the pathogen to escape the central grove and to infect plants in the surrounding groves , although this does happen occasionally . Snapshots from other runs indicate that these aspects of BBSC dynamics are general for ; spread is localised with separate foci of infection that grow and coalesce over time , and spread is also largely restricted to the central grove , at least for the first to years . Varying the initial level of infection indicates the final number of productive ( i . e . asymptomatic ) plants at , , is highly dependent on ( Figure 4 ( a ) ) , at least for low values of . However , since decreases sharply with the amount of inoculum that is initially present , effectively the whole of the central grove becomes infected by for . The value of depends strongly on the planting density ( Figure 4 ( b ) ) , with low host density leading to very little spread and so high values of ( again with ) . However at more realistic planting densities the spread is much more devastating . On average only of plants escape ( visible ) disease by years at the density of the typical grove plants per hectare ) . This behaviour leads to a disease-driven trade-off in the number of productive plants per hectare . Low planting density can give excellent disease control , with very high values of , but of course also implies fewer plants per hectare . The optimum density of productive plants is therefore recovered at an intermediate host spacing: for , this was at a planting density of around 200 plants per hectare , with per hectare ( Figure 4 ( c ) ) . This qualitative result is robust to the initial level of infection , and there was an optimum planting density for all values of we considered . However both the optimal planting density , and per hectare at this planting density , decreased as the initial level of infection was increased ( Figure 4 ( d ) ) , although these responses begin to flatten off for . We also considered the response of the yield ( cf . Equation 6 ) to the planting density . Again for a given level of initial infection , a planting density that leads to an optimum yield per hectare can be defined ( Figure 4 ( e ) ) , although the density that optimises yield when ( plants per hectare ) is larger than that required to maximise the value of ( plants per hectare , as described above ) . The response was also differently shaped , with the yield per hectare remaining at a non-zero value for even very large planting densities ( compare 4 ( c ) with 4 ( e ) ) . This is because even at high densities the epidemic does not infect the entire central grove within the first few years of the epidemic , and so the yield is then non-zero ( see also the inset to Figure 4 ( e ) , which shows the yield before normalisation of to fixed grove area ) . However , the response of the optimum planting density required to optimise yield per hectare for different values of the initial level of infection , and the response of the optimum yield per hectare itself at optimum planting density to the initial level of infection both follow a similar pattern to the responses for ( compare Figure 4 ( d ) and Figure 4 ( f ) ) . Even at relatively high initial levels of infection , , roguing can lead to excellent disease control ( Figure 5 ( a ) ) . At ( a level at which every plant within the central grove would become infected without control within 20 years ) , even the rather long roguing interval would save approximately 20% of plants from visible symptoms at . As is shortened , of course increases . Values of months lead to high levels of disease control ( e . g . % ) , and even gives . This response is comparatively robust to the initial level of infection ( Figure 5 ( b ) ) : although does decrease as is increased ( for fixed ) , it does so only relatively slowly . The value of in fact always depends on and in this broad fashion ( Figure 5 ( c ) ) , decreasing as either parameter is increased . For short roguing intervals , however , was relatively irresponsive to , and indeed there was a large set of pairs for which excellent control was achieved . This was despite the more restricted range of pairs of these parameters for which the pathogen was reliably eradicated from both the central and the surrounding groves ( Figure 5 ( d ) ) . We also examined the response of the median value of to changes in the roguing interval , , and the probabilty of detection , ( Figure 5 ( e ) ) . Unsurprisingly , the impact of the epidemic is increased as is increased or is decreased . In fact the shape of the contours of constant can be explained by a simple calculation . If the other epidemiological parameters are fixed , the efficacy of roguing is set by the effective infectious period of the average host . This is the time for which the host is infectious , i . e . the time between the emergence of infectivity after the latent period has passed and later removal of the host by roguing . If the probability of detection is , then the number of surveys required to detect a host after the emergence of symptoms upon it is a geometric random variable , with average . A particular symptomatic plant could have become infectious at any time between the final round of surveying when it was asymptomatic/uninfectious and subsequent round by which time it was symptomatic . If we assume the time of the transition between states and in our model is uniformly distributed between surveys ( i . e . if we ignore any knock on effect from the slight increase in the rate of infection between rounds of detection that would occur because the number of infected plants increases between surveys ) , then the average effective infectious period can be approximated by ( 10 ) For the default parameters and , the average infectious period is ; all pairs with this effective infectious period are shown by the black curve in Figure 5 ( e ) .
We used Markov chain Monte Carlo with data augmentation to fit a spatially-explicit , stochastic , epidemiological model to the spread of BBSC , and have estimated a number of key epidemiological parameters . Dispersal was exponential , with median approximately 5 m ( similar to the distance between neighbouring pairs of plants in a typical citrus grove in Brazil ) . Laranjeira et al . [20] suggest that the BBSC pathogen may be transmitted by an air-borne vector of limited dispersion ability , and our results are consistent with that possibility . Our estimate of the dispersal scale , together with a careful review of the dispersion ability of arthropods detected in the Bahia region , may help to narrow the set of candidate vectors . Certainly a number of mites and scale insects are known to transmit viral diseases , both in citrus [74] and other perennials [75] , and similar species would be an obvious place to begin such a search . Our parameter estimates are also consistent with an association between a bark wounding insect and a splash dispersed fungus . To obtain an adequate fit to the experimental data we included a delay for plants to reach epidemiological maturity before being able to spread and/or show symptoms of the disease in our model . While it is of course rather difficult to give a mechanistic interpretation of this delay because of the uncertainities surrounding BBSC etiology , it could , for example , correspond to a need for mature tissues for symptom expression , or a bark borer insect vector that only feeds on mature bark . Irrespective of its mechanistic basis , our estimate of the delay is approximately 24 months . Laranjeira et al . [20] took the long delay before disease began to spread in the experiment as indicative of the incubation period for the pathogen that causes BBSC , which we instead estimated to be approximately 6 months . Given the very good statistical support for our model fitting , we contend that our new interpretation of the experimental results is more plausible , especially since a two year incubation period is rather long for a vectored disease . In a grove at planting density typical of citrus production in Brazil , we predict that BBSC would spread slowly . This was unsurprising given the relatively slow rate of disease spread in the original experiment , in which the density of host plants was approximately six times higher than found in citrus production . Nevertheless , and slow spread notwithstanding , we predict BBSC would easily spread throughout an entire grove within 20 years , even for modest levels of initial infection ( ) . In turn this indicates that careful sanitation of new plantings for BBSC symptoms is important . Despite the official programs to foster propagative plants under screenhouses in Bahia , symptomatic “mother” plants are still found [62] , and most nurseries are not kept under screenhouses [76] . This clearly presents a risk , particularly since there is no diagnostic test to identify asymptomatic infected plants . This compelled us to investigate other types of control apart from sanitation . We note that , although high BBSC severity and incidence can be routinely detected in mature commercial groves in Bahia , the incidence of disease is usually quite low at the time of first detection ( HP Santos-Filho , personal communication ) . The particular range we used was therefore intended to account for the full range of values that may occur in practice , given groves at different distances from sources of inoculum and/or with different levels of sanitation before planting . The influence of the initial level of infection on the optima we identify indicates that , for practical implementation , it would be advantageous to perform further experimentation and/or further data-collection to enable to be more precisely quantified . We therefore used our model to examine the effect of host spacing on disease spread . As the density of hosts was increased , so did the level of disease , which of course was expected [26] . However this is particularly unfortunate given recent trends toward higher planting densities in commercial citrus production in Brazil [77] . We therefore examined the trade-off between host density and productivity in the presence of disease by considering the density of plants that escape infection over a 20 year timescale as the host spacing was altered . We found an optimum planting density , at which the reduction in productivity due to planting fewer hosts per hectare was offset by the reduced losses to disease ( cf . Figure 3 ( c ) ) . Although the exact nature of this optimum depended on the initial level of infection , optimal densities were typically sufficiently low that there would be enough space for an intercrop to be established . This approach is already used in Brazil , where growers sometimes plant passion fruit or pineapple between rows of citrus . However , since the intercrop would undoubtedly have its own effect ( s ) on pathogen dispersal [32] , [33] , investigating the epidemiological consequences of intercropping requires more data . According to our simulation results , roguing , even when detection is imperfect , can control disease successfully ( cf . Figure 4 ) . Control can be achieved for relatively long roguing intervals , even for high levels of initial infection . Indeed in our scans showing the effect of roguing interval on control efficacy we used a default value of ( rather than as used in assessing the effect of host density ) in order to obtain a more meaningful response as the parameters of interest were changed . This good level of control was possible because of the slow rate of BBSC spread and its limited dispersal ability . Control by roguing is also aided by the absence of cryptic infection ( i . e . hosts that are able to infect without showing symptoms ) . This contrasts with a number of other pathogens of citrus , for example Xanthomonas axonopodis , the bacterium that causes citrus canker , for which there is both significant long-range dispersal [78] and cryptic infection [6] . Indeed the recent attempt to eradicate citrus canker from Florida involved removing any host plant within 579 . 1 m ( 1900 ft ) of a detected symptomatic focal plant , irrespective of apparent disease status [79] . However , the epidemiology of BBSC indicates that a similar approach is not required here , and initial tests of this type of control strategy indicated that it did not noticably outperform simple roguing ( data not shown ) . Control was possible even though roguing only occurred within the central grove . It did not require the pathogen to be entirely eradicated from the system , and indeed for high values of , the pathogen was eradicated only rarely ( cf . Figure 5 ( d ) ) , presumably because there was at least one escape of the pathogen from the central grove before it was effectively controlled there . This surprisingly high level of control despite an ever-increasing external reservoir reflects the low probability of the pathogen returning to the central grove once it has escaped ( cf . Figure 1 ( c ) ) , and on the occasions it does return , frequent roguing limits its impact . Ultimatately this derives again from the limited disperal ability of the pathogen that causes BBSC . For pathogens capable of faster and/or long-distance dispersal , synchronisation in control is acknowledged to be extremely important , since otherwise the pathogen is able to persist , bulk-up and repeatedly cause devastating reinvasion from uncontrolled areas that act as refugia [10] . Following common practice in the Brazilian citrus industry , removed plants were not replaced in our model , which again facilitated control . Replanting removed trees results in a constant supply of new susceptible hosts to areas with infection , which necessarily makes control more difficult . The efficacy of roguing was characterised by considering , the average effective infectious period ( Equation 10 ) , and this quantity was an excellent predictor of the number of plants that escape disease ( cf . Figure 5 ( e ) ) . Investigating how this result generalises to pathogens that are harder to control would be an interesting extension , particularly because the approximation used in the calculation of is most accurate for pathogens that spread slowly . We note that , although simple , the principle underlying the calculation of has been reported incorrectly in previous studies that used non-spatial , compartmental models . Parameter values given in Table 2 of Jeger et al . [3] ( see also Madden et al . [80] ) indicate that if roguing is performed monthly then the equivalent removal rate would be . This assumes that symptoms and infectivity are developed immediately after rounds of surveys , and so that the average infectious period is . Given the more accurate estimate of 15 days , the rate of removal for monthly surveys with perfect detection should in fact be . By introducing a simple measure of the profitability of a grove , we demonstrated the trade-off between the cost of detection and the benefits of control ( cf . Figure 6 ) . An optimum roguing frequency can be determined , balancing the increased cost of roguing more frequently against the improved control it leads to , although this optimum is conditioned on the initial level of disease ( cf . Figure 6 ( d ) ) and the cost of examining a plant for disease symptoms relative to the difference between the sale price of the fruit from a single year's harvest and the yearly cost of cultivating a tree . For simplicity and ease of presentation , our definition of the cost of control focused exclusively on the cost of detection and did not include the cost of removal . However , because an individual plant would potentially be surveyed many times , but can be removed at most once , we believe this is a reasonable simplification . While our methodology could readily be extended to include more complex economics ( e . g . removal costs , cost of initial grove establishment , increased yield from older plants ) , or to allow for growers potentially ceasing cultivation if the net profit from a particular grove fell below zero despite the yield that would subsequently accrue , the broad result would certainly be robust to these changes . A more interesting extension would be control strategies that change over time . An example of this is a roguing interval that depends on the current ( observed ) prevalence of infection , and so that could cause surveying to slow down or even stop once the disease was judged to be under control . This differs from the implementation considered here , in which the cost of detection for low levels of initial infection and short roguing intervals may be overstated: any grower who surveyed weekly but did not find disease for a number of years would doubtless reduce the frequency of surveying or even stop entirely . Investigating this type of adaptive strategy , together with the consequential risk of failure that derives from having to predict whether the disease has actually been eradicated or has merely not been found recently , will form the basis of our future work in this area . A number of previous models have used deterministic mean field representations of cultural control [1] , [2] , [54] , [55] , [81] . More recently stochastic , spatially-explicit models have predominated [10] , [56]–[58] , although typically these models are not fitted to data ( a series of studies of the failed eradication of citrus canker in Florida are the exception [6]–[8] ) . What previous models lack , however , is a treatment of the economic aspects of control , and the trade-offs and optima to which this can lead . While significant progress in examining this type of trade-off has been made using optimal control theory [59] , [82] , [83] , the complexity of the associated mathematics has necessarily reverted attention to deterministic , non-spatial models . Using a spatial , stochastic model parameterised with real data to balance the benefits of effective disease control against its costs is the novel aspect of our work . In addition to the additional insight into BBSC epidemiology obtained by our model fitting , providing a “real world” example showing how a mathematical model can be used to optimise and test both the epidemiological and economic aspects of control strategies for a plant disease is therefore the key contribution of this paper . | We consider how mathematical models can be used to inform the control of plant disease , even when the identity and biology of the pathogen are not well understood . This is often the case: control of emerging epidemics is most likely to have a significant effect when epidemics remain small , but little may then be known . We analyse data from an experimental plot concerning spread of Bahia bark scaling of citrus , an economically-important disease in north-eastern Brazil , by fitting a mathematical model , which also accounts for uncertainty , to disease spread . Our model captures the epidemiological features of the disease , revealing that transmission is localised and that disease spreads relatively slowly . We use the model to investigate fundamental trade-offs underlying cultural disease control at scales relevant to citrus production . We show how optimal planting densities can be defined , which balance slower spread of disease against the profit that would be lost by growing fewer plants . We also show how the cost of looking for and removing symptomatically diseased plants can be balanced against the reduced disease it leads to . Ours is the first study to consider how a parameterised mathematical model can be used to design optimised cultural controls of plant disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"population",
"modeling",
"biology",
"and",
"life",
"sciences",
"population",
"biology",
"infectious",
"disease",
"modeling",
"computational",
"biology",
"infectious",
"disease",
"control"
] | 2014 | Cost-Effective Control of Plant Disease When Epidemiological Knowledge Is Incomplete: Modelling Bahia Bark Scaling of Citrus |
Dissection of the genetic architecture of complex traits persists as a major challenge in biology; despite considerable efforts , much remains unclear including the role and importance of genetic interactions . This study provides empirical evidence for a strong and persistent contribution of both second- and third-order epistatic interactions to long-term selection response for body weight in two divergently selected chicken lines . We earlier reported a network of interacting loci with large effects on body weight in an F2 intercross between these high– and low–body weight lines . Here , most pair-wise interactions in the network are replicated in an independent eight-generation advanced intercross line ( AIL ) . The original report showed an important contribution of capacitating epistasis to growth , meaning that the genotype at a hub in the network releases the effects of one or several peripheral loci . After fine-mapping of the loci in the AIL , we show that these interactions were persistent over time . The replication of five of six originally reported epistatic loci , as well as the capacitating epistasis , provides strong empirical evidence that the originally observed epistasis is of biological importance and is a contributor in the genetic architecture of this population . The stability of genetic interaction mechanisms over time indicates a non-transient role of epistasis on phenotypic change . Third-order epistasis was for the first time examined in this study and was shown to make an important contribution to growth , which suggests that the genetic architecture of growth is more complex than can be explained by two-locus interactions only . Our results illustrate the importance of designing studies that facilitate exploration of epistasis in populations for obtaining a comprehensive understanding of the genetics underlying a complex trait .
In genomics , gene interaction networks have been evidenced in numerous biological systems using e . g . expression profiling [10] , protein-protein interaction [11] and gene-knockout [12] studies . In genetics , although the theory is established and has been discussed , replicable genetic evidence linking genetic polymorphisms in genetic networks to phenotypic expression has proven difficult to obtain . In this paper , we describe the replication and in-depth exploration of a multi-locus gene-interaction network previously shown to explain nearly half of the long-term selection response in a bi-directional selection experiment in chickens [13] . It was previously shown that this gene-interaction network did not only explain responses to selection , but was also a likely contributor to the smaller than expected decreases in genetic variation in the selected lines [14] as well as the lower than expected power to map individual loci determining body weight ( i . e . the selected trait ) for which the lines showed an eight-fold phenotypic difference [15] . The earlier results were obtained using data from the original finding of the network in an F2 cross between the selected lines [13] . As a result , the resolution of the inferred QTL was limited to confidence intervals covering >10 Mb in each identified locus . This did not allow for discrimination of whether the estimated genetic effect estimates were due to one interacting gene in each of the segments or if they were a composite effect of a larger number of interacting genes . To replicate and study the QTL inferred in the original population , an eight-generation advanced intercross line ( AIL ) was bred from the founder individuals of the original F2 population [13] . These data were recently used by Besnier et al [16] to screen nine selected chromosomal segments for marginal ( additive ) QTL effects , and ten loci in these segments had significant individual effects on body weight . Six of the nine chromosome regions contained loci involved in the radial network of epistatic QTL reported by [13] and the central locus in that network , Growth9 , was shown to contain two independent QTL , which were then designated Growth9 . 1 and Growth9 . 2 . Here , we used the large AIL pedigree to study whether the original finding of strong epistatic interactions in QTL network replicated in this independent dataset and to explore the network further by extending the analyses to also include three-locus interactions . By replicating , fine mapping and extending the QTL-network in this independent population , we show that the epistasis can be stable across generations , i . e . the estimates of the QTL interactions are similar , and that interactions higher than second order are important in the genetic architecture of the selected trait . The implications of the replication of epistatic interactions across generations and novel insights that can be gained from also studying three-locus interactions are discussed . Suggestions are also given concerning designs of future studies to elucidate the role of epistasis in the genetic architecture of complex traits . A new method for performing genome wide scans for third order interactions is also introduced . When a locus acts as a capacitor , its genotype modifies the genetic effects of other loci to be either smaller or larger . For third order epistasis , capacitation can be studied by examining the genotype-phenotype map for triplets of loci , and comparing differences between the “planes” or “slices” of the three dimensional map . Each plane is a two-locus genotype-phenotype map and the effect of the third locus is observed as the distance between the planes for the other two loci ( Figure 1 ) . Let us assume a phenotype governed by three loci - Q1 , Q2 and Q3 - where Q1 is the conditioning locus . If neither locus has any effect , the planes corresponding to the “HH” , and “LL” genotypes for Q1 should be flat across the genotypes of Q2 and Q3 and also at the same level ( Figure 1a ) . If Q1 , and only Q1 , has a marginal effect the planes will still be horizontal , with spatial separation due to the marginal effect ( Figure 1b ) . If all three loci have ( non-interacting ) marginal effects ( Figure 1c ) , or even if Q2 and Q3 interact with each other but not with Q1 ( Figure 1d ) , the planes will no longer be horizontal , but they will have the same shape . If , however , Q1 is interacting with the other two loci , the shape of the planes will differ , and result in them having different within-plane variances ( Figure 1e ) . In particular , if the locus has a capacitating effect , the difference in variance between the planes should be substantial . By measuring the ratio between the plane with the highest variance and the plane with the lowest variance ( henceforth Rp ) the capacitating epistatic effect of each locus in a given triplet can be quantified . Calculating this type of measure is computationally efficient compared to regression-based methods , which opens new possibilities for developing more refined strategies for identifying gene-gene interactions . However , scaling and significance testing is not straightforward , as the range of the ratio is strongly affected by the magnitude of the smallest variance . Screening for an effect of loci on the variance rather than the mean appears to have a large potential for identification of interacting loci and the computational efficiency should make exhaustive scans of large data sets feasible in spite of the high dimensionality .
From [16] , it was known that the nine segments to be examined in the AIL contained loci with either strong or suggestive evidence of marginal effects on body weight . Here , we aim to explore whether any of these loci also display epistatic effects . Of particular interest was to determine if we could replicate the original radial epistatic network around Growth9 [13] , that was later shown in [16] to contain two separate QTL: Growth9 . 1 and Growth9 . 2 , was i ) also present in the AIL and ii ) displayed the same type of capacitating epistasis as in the F2 population . To move beyond the observations related to the original network and understand the combined effect of all the identified interacting QTL , we examined the phenotypic values for individuals with a given set of genotypes for all possible pairs and triplets of loci within a set of six loci; the radial network ( Growth2 , Growth4 , Growth6 , Growth9 and Growth12 ) and the QTL Growth1 . Growth1 was included because it had a strong individual marginal effect , displayed suggestive interactions in the exhaustive 2D-scan ( Figure S1 ) , and had significant interaction effects with four of the other loci when used in a stratified scan of the same type as described for the other loci above ( data not shown ) . Using a large multigenerational pedigree we are able to demonstrate replication of a multi-locus QTL interaction network in vertebrates . Our results show that the type of interactions detected in the original population replicate to a considerable extent , both regarding the loci included as well as their combined effects . Due to the large population size we were also able to include analyses of higher order interactions that show that the primary interaction mechanism , genetic capacitation , is a main feature of the network that involves not only pairs but also triplets of loci . Based on these results we introduce the idea that high order epistasis can be studied by examining the variance differences between genotypes in multi-dimensional genotype-phenotype maps . This study provides further evidence for the importance of genetic interactions in determining complex phenotypes and indicates that the value of epistatic analyses in studies aiming at genetic dissection of the architecture of complex traits .
An eight generation Advanced intercross line ( AIL ) was produced from two selected lines of chickens obtained by bi-directional , single trait selection for bodyweight at 56 days of age ( referred to as the High Weight Selected “HWS” and Low Weight Selected “LWS” lines ) . The lines originate from a common base population , consisting of crosses of seven partially inbred lines of White Plymouth Rock chickens [22] , [23] . All procedures involving animals used in this experiment were carried out in accordance with the Virginia Tech Animal Care Committee animal use protocols . Individuals from generation 40 of the HWS and LWS lines were used as founders for the AIL . The sex-averaged 56-day body weight at this generation was 1522 g ( SE: ±36 g ) for the HWS line and 181 g ( SE: ±5 g ) for the LWS line . The observed mean heterozygosity , H0 , at all autosomal loci was calculated as 0 . 146 and 0 . 156 in the high and low lines , respectively [24] . The husbandry of the intercross was identical to that of [19] . To produce 100 F1 progeny , 10 HWS males were mated with 22 LWS females and 8 LWS males were mated to 19 HWS females ) . About 100 individuals were produced in generations F2 , F4 , F5 , F6 and F7 and 300 and 400 individuals in generation F3 and F8 respectively . The F8 generation has undergone the most number of recombination events and should therefore give the best resolution , which is why it is also the largest . The f3 generations was increased in number in order to provide power to detect variation that might be lost deeper in the pedigree . Nine chromosome regions with significant or suggestive QTL for body weight in the F2 generation [13] , [15] , [20] were selected for further study in the AIL . The segments are abbreviated as in [15] . For all individuals in the AIL , DNA was extracted from blood by AGOWA GmbH ( Berlin , Germany ) . In addition , 15 individuals from each parental line were genotyped for approximately 13 , 000 SNP markers , distributed genome-wide , as described in [15] . The nine QTL regions include 384 segregating SNPs , selected out of the 13 , 000 total SNPs . The average distance between the markers was less than 1 cM . All individuals in the AIL ( n = 1529 ) were genotyped for the set of 384 markers using the GoldenGate assay ( Illumina , CA ) at the SNP technology platform in Uppsala ( Sweden ) . To study the effects of the radial network we used the stratification based QTL interaction analysis employed in [13] . A brief description of the procedure is provided below . First , QTL genotype contrasts ( regression coefficients that are essentially re-scaled probabilities [25] ) were estimated at each marker in the 10 genomic regions evaluated . These contrasts , calculated from a gametic IBD matrix [26]–[29] , were used in a two-dimensional scan for interacting QTL pairs where tests were performed using a two-locus epistatic model [17] . Several types ( single QTL at each locus , two non-interacting QTL and two interacting QTL ) of models were fitted and the loci were considered to be interacting when the interacting QTL model had a significant improvement over the others . Significance was determined from a permutation test procedure , as described in [17] . Based on this analysis , we selected the marker location in each segment that showed the strongest support for interactions as the locus to be tested for significant epistasis . We generated a series of stratified datasets , i . e . subsets of the data that contained only the individual most likely to be homozygous at a conditioning locus . Two strata for each conditioning locus ( one stratum with the LL homozygotes and one with the HH homozygotes ) was produced using the genotypic contrasts at the marker position in each selected chromosomal segment that had the strongest indication of epistasis in the two-dimensional QTL scan . We stratified using six conditioning loci ( Growth1 , Growth2 , Growth4 , Growth6 , Growth9 and Growth12 ) , yielding a total of 12 subsets . The sizes of the strata varied between 229 and 377 individuals . These stratified datasets were then used for the interaction analyses . For pairs of loci where significant interactions were detected , the additive genetic effects of the QTL were estimated separately in each of the three strata . The estimates of the additive effects were obtained using linear regression of residual body weight at 56 days of age , corrected for the fixed effect of generation and sex , as in [13] . The robustness of potential interactions was evaluated using a combined bootstrap and permutation testing strategy , designed to control the type-I errors in this advanced intercross pedigree [30] . For each locus tested , the same analyses were performed for the whole data set , as well as for the HH and LL strata separately . In each dataset , 200 bootstrap samples were generated and analyzed using a one-dimensional QTL scan [31] . For each marker , an additive model was fitted , with sex and generation as fixed effects , the results averaged over all bootstrap replicates . Bootstrap data sets were sampled with replacement and were of the same size as the original data set . The significance threshold was determined by a permutation test , where 1000 datasets were generated by permuting the genotypes of the individuals , while preserving the relationship between phenotype and the two fixed factors , sex and generation in the pedigree . The threshold was selected so that it was above the maximal value found in 95% of the permutations . One-dimensional scans , following the procedure described above , were performed in each dataset , and the distribution of the maximal F-scores from each replicate was used to determine an empirical significance threshold . Loci defined as epistatic in the two-dimensional scan were included in a joint network analysis as follows: initially , all marginal and two-way interactions for epistatic loci in the network were estimated jointly using the NOIA framework [18] . The use of the “statistical model” in NOIA provided an orthogonal model for the estimation of genetic effects , even though the population did not conform to an ideal F2 allelic frequency distribution at each locus . From the orthogonal estimates , a multi-locus genotype-phenotype ( GP ) map was constructed using the “transformation” operation in NOIA [18] . This GP-map provides estimates of the genotype values ( expected phenotypes ) for all multi-locus genotype combinations in the network that are useful for functional studies of epistatic interactions . In addition , phenotypic means of genotype classes were estimated directly using a discretized version of the data set . Genotypes of each individual were discretized at the marker positions based on the genotype contrast value . The contrasts ranged from −1 to 1 , with 0 . 4/−0 . 4 used as thresholds for assigning a HH , HL , or LL genotype to a marker . These thresholds are conservative ( the HL interval is larger than that of the homozygotes; 0 . 33/−033 would be most natural and split the space evenly ) , to reduce the number of falsely assigned homozygotes because they are most critical to the outcome . However , because most individuals had accurately estimated genotype probabilities , i . e . contrast values close to −1 , 0 or 1 , the choice of cut off has little effect on the outcome . We also used the discretized data in a second order interaction model-based NOIA-approach , and constructed a partial GP map directly from the phenotypic means of individuals with given genotypes . The observed means of the genotype classes were then compared to the expected ones from the NOIA model . Differences between these values are an indication of the effect of higher order interactions . Discretized data were used to study three-way interactions . For a set of triplets of loci , the phenotypic mean ( again corrected for sex and generation ) for each three-locus genotype was calculated . Thus , for each triplet of loci , means were calculated for 27 classes of individuals , each corresponding to a three-locus genotype . Since each genotype was scored separately , any type of interaction was detected from the pattern of these values . For detection of capacitating epistasis , the 27 means were grouped into three “planes , ” where each plane consists of the nine genotypes that shared a single genotype of the potentially capacitating locus ( i . e . there is one plane each for the “HH , ” “HL” and “LL” genotypes at that locus ) . Then , the variance within each plane was calculated together with the ratio between the planes with highest and lowest variance . This ratio measures the capacitating effect of the conditioning locus . | This study provides evidence for a strong and persistent contribution of epistatic interactions to the selection response in two chicken lines subjected to 50 generations of divergent selection for 8-week body weight . We show that the genetic architecture of the trait involves genetic interactions of both second- and third-order and that , together , they explain a large portion of the phenotypic divergence between the lines . By replicating a radial epistatic network found in an independent intercross from the same founder individuals , we show that the type of genetic interactions affecting this complex trait is persistent over time . In addition to replicating pair-wise interactions , the size of the pedigree also facilitated evaluation of third-order interactions , which allowed us to further describe the complex genetic mechanisms underlying growth phenotype in chicken . Moreover , a new approach for measuring and detecting capacitating epistasis was proposed . By showing the importance of third-order epistasis in this system , we reinforce the importance of taking it into account when designing experiments aimed at elucidating the genetic architecture of complex traits . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"genetics",
"genetic",
"networks",
"genome",
"evolution",
"genome",
"scans",
"chicken",
"statistics",
"quantitative",
"traits",
"epistasis",
"animal",
"models",
"genome",
"analysis",
"tools",
"model",
"organisms",
"mathematics",
"trait",
"locus",
"analysis",
... | 2011 | Replication and Explorations of High-Order Epistasis Using a Large Advanced Intercross Line Pedigree |
Frataxin ( Yfh1 in yeast ) is a conserved protein and deficiency leads to the neurodegenerative disease Friedreich’s ataxia . Frataxin is a critical protein for Fe-S cluster assembly in mitochondria , interacting with other components of the Fe-S cluster machinery , including cysteine desulfurase Nfs1 , Isd11 and the Isu1 scaffold protein . Yeast Isu1 with the methionine to isoleucine substitution ( M141I ) , in which the E . coli amino acid is inserted at this position , corrected most of the phenotypes that result from lack of Yfh1 in yeast . This suppressor Isu1 behaved as a genetic dominant . Furthermore frataxin-bypass activity required a completely functional Nfs1 and correlated with the presence of efficient scaffold function . A screen of random Isu1 mutations for frataxin-bypass activity identified only M141 substitutions , including Ile , Cys , Leu , or Val . In each case , mitochondrial Nfs1 persulfide formation was enhanced , and mitochondrial Fe-S cluster assembly was improved in the absence of frataxin . Direct targeting of the entire E . coli IscU to ∆yfh1 mitochondria also ameliorated the mutant phenotypes . In contrast , expression of IscU with the reverse substitution i . e . IscU with Ile to Met change led to worsening of the ∆yfh1 phenotypes , including severely compromised growth , increased sensitivity to oxygen , deficiency in Fe-S clusters and heme , and impaired iron homeostasis . A bioinformatic survey of eukaryotic Isu1/prokaryotic IscU database entries sorted on the amino acid utilized at the M141 position identified unique groupings , with virtually all of the eukaryotic scaffolds using Met , and the preponderance of prokaryotic scaffolds using other amino acids . The frataxin-bypassing amino acids Cys , Ile , Leu , or Val , were found predominantly in prokaryotes . This amino acid position 141 is unique in Isu1 , and the frataxin-bypass effect likely mimics a conserved and ancient feature of the prokaryotic Fe-S cluster assembly machinery .
Frataxin is a highly conserved protein that is found in both prokaryotic and eukaryotic organisms [1] . The protein was originally identified based on its connection to Friedreich’s ataxia , which is an inherited neurodegenerative and cardiodegenerative disease resulting from a deficiency of frataxin [2] . Recently a mitochondrial Fe-S cluster assembly protein complex was identified consisting of frataxin in association with the cysteine desulfurase Nfs1 , the small eukaryote-specific protein Isd11 , and the scaffold protein Isu1 [3 , 4 , 5] . This protein complex serves to synthesize Fe-S cluster intermediates on Isu1 for subsequent transfer to the myriad proteins that use Fe-S cluster cofactors [6 , 7] . Iron-sulfur cluster intermediates contain iron and sulfur bound to the Isu1 scaffold in a [Fe2S2] configuration [8] . Although the precise function of frataxin has not been defined , it probably plays a role in sulfur and/or iron donation to Fe-S cluster intermediates [9 , 10] . The entire process of Fe-S cluster biogenesis is highly conserved between eukaryotic mitochondria and prokaryotic organisms [6 , 7] . Similar components are present , and in many cases the corresponding proteins function as orthologs . Sulfur for Fe-S cluster synthesis is derived from cysteine via the action of a cysteine desulfurase ( Nfs1 in yeast , IscS in bacteria ) . This enzyme binds the amino acid cysteine in a substrate-binding site via the pyridoxal phosphate ( PLP ) cofactor [8] . The bound substrate is subjected to nucleophilic attack by an active site cysteine present on a moveable loop of the protein , forming a persulfide ( e . g . Nfs1-S-SH in yeast ) . The persulfide sulfur is then transferred to recipients including Isu1 and used in building Fe-S clusters [11] . Frataxin ( Yfh1 in yeast , CyaY in E . coli ) interacts with the cysteine desulfurase and may be involved in regulating the enzyme activity . Whereas a positive regulatory effect has been observed for the yeast or human proteins [12 , 13] , a negative regulatory effect has been observed for the E . coli homolog [14] , and this difference has still not been explained [15] . Isd11 is a small accessory subunit that interacts with the eukaryotic Nfs1 and is necessary for its cysteine desulfurase activity [16] . However , Isd11 is eukaryote specific , being entirely absent from prokaryotic lineages [17] . Iron combines with sulfur on the scaffold protein to form Fe-S cluster intermediates . The scaffolds ( Isu1 in yeast , IscU in E . coli ) are highly homologous proteins [18] . The iron donation step is poorly characterized , and frataxin has also been implicated in this step . Both yeast and E . coli frataxins bind iron with low affinity on acidic residues in vitro and interact with their respective scaffold proteins in vitro and in vivo , and thus they may participate in iron donation [19 , 20] . Electrons provided by ferredoxin ( Yah1 in yeast , Fdx in E . coli ) are needed for reduction of iron or sulfur during Fe-S cluster intermediate formation [21] . Following formation of the intermediate on Isu1 or IscU , coordinated by three critical cysteines in the protein backbone , Hsp70 chaperones and cochaperones ( Ssq1 and Jac1 in yeast , HscA and HscB in E . coli ) interact with the scaffolds , mediating Fe-S cluster transfer in an ATP-dependent manner [22] . In terms of phenotypes resulting from frataxin deficiency , however , eukaryotes and prokaryotes show major differences . Total lack of frataxin is lethal in humans and metazoans [4] . Deletion of the YFH1 gene in yeast is associated with extremely deleterious effects , including slow growth , oxidant sensitivity , heme deficiency and lack of Fe-S clusters [23 , 24] . In addition , frataxin deficiency is associated with a curious iron homeostatic phenotype characterized by constitutive and unregulated cellular iron uptake . Within the cell iron accumulates in mitochondria in the form of biologically unavailable ferric phosphate nanoparticles . This constellation of findings apparently results from defective Fe-S proteins in the iron-sensing machinery [25 , 26] . In contrast to the yeast mutants , the effects of frataxin deletion in E . coli are mild . The bacterial deletion strain shows normal growth and does not exhibit iron homeostatic abnormalities or sensitivity to oxidative stress , although in one report the protein level for respiratory complex I was reduced [27] . A spontaneously occurring mutation in a frataxin-deleted yeast strain was found to effectively bypass the severe Δyfh1 phenotypes , restoring normal growth , Fe-S cluster protein levels , iron homeostasis , heme synthesis , and oxidative stress resistance . The effect was conferred by the Met to Ile change of amino acid 141 in the scaffold protein Isu1 [28] . The altered Isu1 was able to bind and activate the Nfs1 cysteine desulfurase in the absence of frataxin , thus providing a possible explanation for the bypass activity [29 , 30] . Interestingly , isoleucine is the amino acid utilized by E . coli in the homologous position of IscU . Thus in yeast lacking frataxin , the Met to Ile change in Isu1 , by substituting the E . coli amino acid at this position , effectively rendered yeast more frataxin independent and more "prokaryote like" . Here we have delved more into the genetics of this frataxin-bypass phenomenon , finding more prokaryotic features of Isu1 bypass mutants . Randomly selected Isu1 bypass mutants were confined to a single amino acid position , and the amino acids conferring bypass were all present in homologous prokaryotic proteins . The prokaryotic homologs were identified both in organisms with frataxin and in organisms without frataxin , underscoring the frataxin-independence associated with these particular scaffold mutants . The entire E . coli IscU ( targeted to mitochondria with a leader sequence ) conferred more frataxin-independence , whereas a reverse substitution , in which the eukaryotic amino acid Met was introduced at the same position , conferred more frataxin-dependence . Examination of the set of Isu1/IscU sequences in available databases also suggests that Isu1-Met appears almost exclusively in eukaryotes and likely coevolved with frataxin . The substituted Isu1-Ile probably mimics a conserved and ancient feature of the prokaryotic Fe-S cluster assembly complex , relieving the frataxin requirement .
The frataxin-bypass mutant was isolated as a spontaneously arising clone of more rapidly growing cells in a background of a frataxin-deleted haploid yeast strain ( Δyfh1 ) [28] . The bypass activity was conferred by a single point mutation ( ATG to ATA ) , changing the amino acid of codon 141 of ISU1 from Met to Ile . However yeast S . cerevisiae carries two redundant genes coding for Fe-S cluster scaffolds , ISU1 and ISU2 [31] . The initial presentation of the suppressor or bypass mutant suggested that it was genetically dominant , because it was able to bypass Δyfh1 even in the presence of a normal copy of ISU2 . In order to evaluate this genetic dominance further , matched Δyfh1 strains were compared , one expressing a single copy of substituted ISU1 and deleted ISU2 , called Δyfh1 [ISU1-Ile] ( Table 1 ) and another expressing both ISU1 and ISU2 in addition to the plasmid-borne substituted Isu1-M141I , called Δyfh1 [ISU1-Ile] ISU1 ISU2 ( Table 1 ) . As assessed by colony formation , both strains showed improved growth compared with the Δyfh1 control ( Fig 1A , compare rows 3 and 4 to row 2 ) , although the single copy [ISU1-Ile] showed slightly better Δyfh1 suppression activity as assessed by growth ( Fig 1A , compare rows 3 and 4 ) , and other phenotypes such as iron uptake and Fe-S cluster levels . Thus , a high degree of genetic dominance of the Isu1 Met to Ile substitution was observed in producing reversal of Δyfh1 phenotypes . ISU1 and ISU2 encode highly homologous proteins , with 83% amino acid identity . Isu1 and Isu2 proteins are functionally redundant , although the endogenous expression level of Isu1 is roughly seven times higher than that of Isu2 , and Isu1 represents most of the cellular Isu protein in a wild-type strain [32] . Significantly the critical Met ( amino acid 141 in Isu1 or 133 in Isu2 ) is present in both proteins . The adjacent cysteine that functions as an Fe-S cluster ligand and the adjoining frataxin-binding motif are also present in both proteins [33] . ISU2 and ISU2-Ile , in which the Met-133 was changed to Ile , were experimentally evaluated . Results show that ISU2-Ile conferred frataxin-bypass activity in a YFH1 shuffle strain ( S1 Fig ) . Bypass activity was conferred whether it was expressed from the native ISU2 promoter or from the ISU1 promoter ( S1 Fig ) . Thus , genetic dominance for frataxin-bypass by ISU2-Ile was observed , similar to that of ISU1-Ile . Furthermore , the ISU2 and ISU2-Ile constructs were also able to support growth of the GAL1-ISU1/Δisu2 strain indicating that they were functional as Isu . The reason that only the ISU1-Ile was isolated and ISU2-Ile was not isolated in the original screen for Δyfh1 suppressors is probably due to the lack of saturation of this genetic screen . All 20 possible amino acids were substituted at position M141 in a single copy plasmid-borne ISU1 . Each mutant form of ISU1 was tested for frataxin-bypassing activity by transforming the plasmids into a YFH1 shuffle strain , followed by counterselection with fluoroorotic acid on raffinose medium to remove the covering plasmid ( Fig 1B , plates 1–3 ) . After counterselection , robust growth was noted for positive control YCplac22-YFH1 , and slow growth was noted for the empty plasmid , YCplac22 ( Fig 1B , plate 1 , “YFH1” versus “-” ) , consistent with the important role of frataxin for normal growth . The ISU1-Ile plasmid with the Met to Ile change restored robust growth in the absence of YFH1 ( Fig 1B , plate 1 , “I” ) . Similarly , plasmids substituted at the same amino acid position and containing ISU1-Leu , ISU1-Cys , and ISU1-Val , conferred improved growth ( Fig 1B , plates 1–3 ) . These ISU1 alleles thus carried frataxin-bypass activity and were genetically dominant , given that wild-type genomic copies of ISU1 and ISU2 were still present . Fe-S cluster assembly scaffold function was tested separately . The collection of ISU1 alleles was transformed into the GAL1-ISU1/Δisu2 strain . In this strain the redundant ISU2 was deleted and ISU1 was placed under control of GAL1 , a galactose-dependent promoter . When the transformants were shifted to a non-inducing carbon source ( i . e . raffinose-based medium ) , only plasmid-borne ISU1 was expressed , allowing scaffold function for each allele to be scored . The wild-type ISU1 ( Fig 1B , plate 4 , M ) supported normal growth , and a large set of ISU1 plasmids with amino acid substitutions also supported normal growth , indicating that these were highly functional ISU1 proteins ( Fig 1B , plates 4–6 , I , L , A , N , C , F , Y , S , V , G , T and H ) . Importantly , all of the substitutions conferring frataxin-bypass activity ( Fig 1B , plates 1–3 , I , L , C , and V ) were also functional ISU1 scaffold proteins . Another set of substitutions was partially functional as shown by slowed growth ( Fig 1B , plate 5 , D , E and W ) , and these mutants tended to accumulate iron , indicating that they were hypomorphic mutants in the Fe-S cluster assembly pathway . A small set of substituted alleles was completely non-functional and did not grow at all on the raffinose plates ( Fig 1B , plates 4–6 , R , Q , K and P ) . The E . coli ortholog , IscU , was studied by NMR methodology [34] and found to assume interconverting disordered or structured conformations . When a conserved amino acid in the primary sequence , Asn 123 in the Isu1 numbering , was substituted with Asp , the conformation of the mutant protein was shown to be shifted preferentially to the disordered form . Alternatively when Asn 123 was substituted with Ala , the conformation was shifted to the structured form [34] . We wondered if the frataxin-bypassing activity could be related to one or the other of these conformations . The Asp change ( N-D; presumably disordered ) and the Ala change ( N-A; presumably structured ) were introduced into YCplac22-ISU1 and the respective plasmids were transformed into the YFH1 shuffle strain . However , neither conferred any bypass activity ( Fig 1B , plate 3 , N-D and N-A ) . The Asp 123 form conferred growth to the GAL1-ISU1/Δisu2 strain indicating efficient scaffold function , but the Ala 123 form was poorly functional in this assay ( Fig 1B , plate 6 , N-D versus N-A ) . Isu1 with the Met to Ile substitution , ISU1-Ile , exhibited both frataxin-bypassing activity and scaffold activity . Therefore the question arose of whether both activities must necessarily reside in the same molecule . Alternatively , the ISU1-Ile might act on wild-type ISU1 protein , stimulating it to perform as the primary scaffold . The existence of a protein complex containing the suppressor ISU1-Ile protein and the wild-type ISU1 protein could provide a physical context for such stimulation to occur [35 , 36] . To begin to address this question genetically , the three cysteine residues of Isu1 , presumed to be Fe-S cluster ligands [31] , were individually replaced by alanine . The C69A , C96A and C139A forms of Isu1 were introduced into strain GAL1-ISU1/Δisu2 , and they were unable to support growth on their own in non-inducing glucose-containing medium . Double mutants were then constructed in which the critical cysteine substitutions were combined with the suppressor ISU1-Ile change of residue 141 in the same molecule . The mutated ISU1 alleles , C69A-M141I , C96A-M141I and C139A-M141I , were tested for frataxin-bypass activity in the YFH1 shuffle strain , and no bypass activity was observed ( Fig 1C , plate 2 ) . Likewise the mutated ISU1 alleles were tested for scaffold activity by expression in the GAL1-ISU1/Δisu2 strain in glucose , and no complementation was observed ( Fig 1C , plate 4 ) . The frataxin-bypass activity of the Cys substitution of residue 141 raised the possibility that the cysteine at this location was functioning as an alternative Fe-S cluster ligand instead of Cys 139 . In a genetic experiment , Cys 139 was replaced with alanine in the presence of the M141C substitution . However , in this case , neither scaffold activity nor bypass activity was observed ( Fig 1C , plates 2 and 4 , number 7 ) , making it unlikely that the Cys substituted at position 141 can function as an alternative Fe-S cluster ligand . In summary , substitutions that abrogated Isu1 scaffold function by altering the Fe-S cluster ligands also abrogated frataxin-bypass function . A mutant allele of NFS1 , nfs1-14 , was identified because of its effects on iron homeostasis , and it was later found to be associated with decreased cysteine desulfurase activity in mitochondria [16] . The cells carrying this mutant allele were viable and grew well in most media despite the low activity of cysteine desulfurase . However , the combination of nfs1-14 and Δyfh1 was synthetically lethal ( Fig 1D , YCp ) . YFH1 ( Fig 1D , YCp-YFH1 ) introduced into the nfs1-14 Δyfh1 strain restored growth . However , although ISU1-Ile was able to bypass Δyfh1 alone , it was unable to rescue the nfs1-14 Δyfh1 double mutant ( Fig 1D , YCp-ISU1-Ile ) . These data suggest that a threshold level of cysteine desulfurase activity is required for the frataxin-bypass activity of ISU1-Ile . Recent biochemical data indicate that frataxin stimulates the cysteine desulfurase activity of Nfs1 [13] . Furthermore , the suppressor ISU1-Ile acts in a similar fashion to stimulate Nfs1 even in the absence of frataxin , perhaps explaining its bypass activity [29] . Thus it makes sense that in the absence of a sufficiently active Nfs1 , bypass of Δyfh1 by the Ile-substituted ISU1 will not occur . The frataxin-bypassing activity of ISU1 protein with changes of Met 141 to Cys , Ile , Val or Leu was shown by growth enhancement of the YFH1 shuffle strain . In this strain , the chromosomal ISU1 and ISU2 were still intact ( Fig 1B ) . To examine and compare these effects in more detail , matched strains were constructed in which each of the bypassing ISU1 alleles was expressed in the absence ISU2 ( Table 1 ) . The starting strain was designated YFH1 [ISU1] , indicating a YFH1 positive strain with plasmid carrying wild-type ISU1 ( Met at position 141 ) ( Table 1 ) . Another strain was designated Δyfh1 [ISU1] , indicating a strain deleted for YFH1 and carrying wild-type ISU1 . Similarly , Δyfh1 [ISU1-Cys] , Δyfh1 [ISU1-Ile] , Δyfh1 [ISU1-Leu] , and Δyfh1 [ISU1-Val] , were used to designate deletions of YFH1 with the indicated ISU1 mutant alleles ( Table 1 ) . The growth of the YFH1 [ISU1] strain was more rapid than the Δyfh1 [ISU1] strain as shown by the larger colony size ( Fig 2A , upper panel , rows 1 and 2 ) . Hydrogen peroxide exposure exacerbated the Δyfh1 phenotype [37] , further slowing growth and viability as a consequence of increased oxidative stress ( Fig 2A , lower panel , rows 1 and 2 ) . The ISU1 alleles with Cys , Leu , or Val at position 141 , similar to the ISU1 M141I allele , improved growth of Δyfh1 under these conditions , including in the presence of hydrogen peroxide , although not quite to levels in the frataxin plus strain ( Fig 2A , upper and lower panels , rows 1–6 ) . In terms of mitochondrial proteins , immunoblotting with anti-frataxin antibody confirmed the correctness of the genetic assignments: the YFH1 strain expressed frataxin ( Fig 2B , Y-M , lane 1 ) and the Δyfh1 strains did not ( Fig 2B , lanes 2–6 ) . The ISU1 alleles were all expressed in mitochondria . The abundance of Isu1 protein was increased in the Δyfh1 [ISU1] strain in the presence of the ISU1-Met allele ( Fig 2B , Δ-M , lane 2 ) , but in the presence of the bypassing ISU1 alleles , Isu1 expression returned to normal , similar to the YFH1 positive strain ( Fig 2B , lanes 3–6 ) . The changes in protein abundance were several fold , and most likely attributed to Aft1/2 transcriptional effects and protein stability effects [32] . Nfs1 protein levels were unchanged in all the strains consistent with the lack of regulation of the protein level ( Fig 2B ) , although cysteine desulfurase activity was strongly regulated ( see below ) . Cytochrome c , a heme protein , was undetectable in the Δyfh1 [ISU1] strain ( Fig 2B , lane 2 ) compared with the wild-type ( Fig 2B , lane 1 ) . The various suppressor strains recovered significant levels of cytochrome c ( Fig 2B , lanes 3–6 ) . Cytochrome c is regulated transcriptionally and post-transcriptionally by heme availability [38] . The dramatic changes in protein abundance probably reflect changes in heme synthesis and steady state heme levels . Iron homeostasis was examined using steady state labeling of growing cells with radioactive 55Fe . In YFH1 [ISU1] strain , cellular iron levels were appropriately regulated , whereas in the mutant Δyfh1 [ISU1] strain , by comparison , excess iron accumulated ( Fig 2C , Cellular iron , Y-M versus Δ-M ) . The presence of the suppressor ISU1 mutant alleles , ISU1-Cys , ISU1-Ile , ISU1-Leu , or ISU1-Val , restored cellular iron levels towards normal ( Fig 2C , Cellular iron , Δ-C , Δ-I , Δ-L , Δ-V ) . The radiolabeled cells were subjected to subcellular fractionation , separating mitochondrial and post-mitochondrial fractions ( Fig 2C , Mitochondrial iron and Post-mito supernatant iron ) . The iron quantitation for these fractions resembled the patterns for whole cell iron . The radiolabeled mitochondrial iron showed features that were dependent on YFH1 . In the YFH1 [ISU1] mitochondria , the predominant portion was solubilized by exposure to non-ionic detergents such as Triton X-100 , whereas in the Δyfh1 [ISU1] mitochondria , proportionally more iron was recovered in the pellet following centrifugation in the presence of detergent ( Fig 2C , Mitochondrial iron distribution , Y-M versus Δ-M ) . Most likely this effect reflects the accumulation of nanoparticles of ferric phosphate that is a hallmark feature of Δyfh1 mitochondria [25] . In mitochondria from the different suppressor strains the amount of insoluble iron was decreased compared with the control deletion strain Δyfh1 [ISU1] , indicating an improvement in the biochemical properties of mitochondrial iron pools ( Fig 2C , Mitochondrial iron distribution , Δ-M versus Δ-C , Δ-I , Δ-L , Δ-V ) . Aconitase ( Aco1 ) , an abundant [Fe4S4] protein of mitochondria , was evaluated . An in- gel assay showed the aconitase enzyme activities in mitochondrial lysates . The control strain YFH1 [ISU1] ( Fig 3A , upper panel , lanes 1 and 7 , Y-M ) showed concentration dependent activity . The frataxin minus strain Δyfh1 [ISU1] ( Fig 3A , upper panel , lanes 2 and 8 , Δ-M ) had no detectable activity regardless of the concentration of lysate in the assay . The various suppressors ( Fig 3A , upper panel , Δ-C , Δ-I , Δ-L , Δ-V , lanes 3–6 and lanes 9–12 ) recovered significant activity , although not entirely to wild-type levels . Aconitase protein ( Fig 3A , lower panel ) as detected by immunoblotting was present in YFH1 [ISU1] control ( lanes 1 and 7 ) , and was decreased in the frataxin-minus Δyfh1 [ISU1] ( lanes 2 and 8 ) mitochondria , perhaps because the apoprotein was turned over more rapidly by mitochondrial proteases . Aconitase protein levels were restored in the suppressor strains ( Fig 3A , lower panel , lanes 3–6 and lanes 9–12 ) , consistent with recovery of enzyme activity . The persulfide-forming activity in these mitochondria was tested as an indication of their cysteine desulfurase activity . Isolated and intact mitochondria were depleted of endogenous nucleotides and NADH by incubation at 30°C for 10 min , thereby blocking Fe-S cluster biogenesis without disrupting cysteine desulfurase activity [16] . After labeling with 35S-cysteine , mitochondrial proteins were separated by non-reducing SDS-PAGE , and the persulfide covalently bound to Nfs1 was visualized by autoradiography ( Fig 3B , Nfs1-S-35SH ) . The signal was absent in nfs1-14 mitochondria , which were unable to form significant Nfs1 persulfide because of a hypomorphic mutation in Nfs1 ( Fig 3B , lane 13 ) [16] . The specificity of the Nfs1-persulfide signal was further confirmed by immunoprecipitation with anti-Nfs1 antibody [16] . Several other radiolabeled bands were detected in mitochondria ( Fig 3B ) , some of which were present in the nfs1-14 control and some of which were absent . Thus these background bands could be attributed to direct binding of 35S-cysteine , persulfide transfer from Nfs1 which was incompletely blocked , or binding of other reactive cysteine persulfides to mitochondrial proteins . The YFH1 [ISU1] mitochondria ( Fig 3B , lanes 1 and 2 , Y-M ) had significantly more Nfs1 persulfide than the frataxin-minus mutant Δyfh1 [ISU1] ( Fig 3B , lanes 3 and 4 , Δ-M ) . Each of the suppressor mutants ( Fig 3B , Δ-C , Δ-I , Δ-L , and Δ-V; lanes 5–12 ) recovered persulfide-forming activity in mitochondria . Differences among the suppressor alleles were not apparent , and each one provided rescue of the persulfide-forming activity that was deficient in Δyfh1 [ISU1] mitochondria . Next , 35S-cysteine labeling of isolated intact mitochondria was performed in the presence of added ATP , GTP , NADH and iron , thereby permitting multiple cycles of Fe-S cluster formation to occur [16] . Soluble proteins from these mitochondria were separated on native gels , and autoradiography was used to detect newly synthesized Fe-S clusters on aconitase ( Fig 3C , Aco1 [Fe-35S] ) . In the”wild-type” YFH1 [ISU1] mitochondria ( Fig 3C , lanes 1 and 2 , Y-M ) a strong signal was observed , whereas in the frataxin-minus Δyfh1 [ISU1] ( Fig 3C , lanes 3 and 4 , Δ-M ) no signal was present , consistent with the important role of frataxin in mitochondrial Fe-S cluster assembly . In the various suppressor mutants , although they still lacked frataxin , Fe-S cluster synthesis on Aco1 was restored to 24–31% of the YFH1 control as assessed by densitometry and correction for the total amount of protein loaded ( Fig 3C , lanes 5–12 ) . A library of randomly mutated ISU1 plasmids was generated by error-prone PCR and transformed into a YFH1 shuffle strain also deleted for ISU1 . The colonies appeared uniform ( Fig 4A , plate 1 ) . The diversity of this library was confirmed by sequencing the inserts from randomly selected colonies . Of 42 colonies evaluated in this way , inserts were identified that included 72 amino acid changes in Isu1 , which were well distributed throughout the coding region ( S1 Table , controls ) . Transformants were replicated to cycloheximide plates , counterselecting against the YFH1-containing plasmid and uncovering the Δyfh1 phenotype . Most colonies grew slowly on glucose or not at all on raffinose , but a few colonies exhibited robust growth ( Fig 4A , plates 2 and 3 ) . Plasmid DNA rescued from these more rapidly growing Δyfh1 colonies was sequenced and in most cases was found to contain single nucleotide changes in ISU1 conferring substitutions of residue 141 of the coding sequence . In some cases , YFH1 sequences were found to have recombined into the ISU1 plasmid and these clones were discarded . The PCR randomization was then repeated with different ISU1 templates starting with Y141 , H141 or F141 , and a large number of colonies ( approximately 36 , 950 representing 17 , 588 amino acid changes in Isu1 ) was screened in this way . The “hits” with frataxin-bypass activity included amino acid changes of M141 to Ile , Cys , Leu , and Val , sometimes in combination with other amino acids changes and sometimes alone ( Fig 4B and S1 Table ) . However , no other amino acid change or combination of changes was able to confer frataxin-bypass activity . All possible amino acid changes in ISU1 were not sampled , and only single nucleotide changes at position 141 were selected . The failure to find amino acid substitutions with bypass activity at other locations in the Isu1 protein may derive from the many constraints on this essential scaffold protein , which must interact with multiple partner proteins and perform multiple functions , such as stimulating Nfs1 activity , coordinating Fe-S clusters and transferring Fe-S clusters [39] . Based on these results , the possibility of finding another ISU1 mutant with frataxin-bypass activity , while not entirely ruled out , seems unlikely . The Met to Ile substitution in yeast ISU1 that conferred frataxin-bypass activity did so by altering the methionine at position 141 ( 107 in the signal-cleaved mature protein ) to the amino acid isoleucine used in the E . coli IscU in the corresponding position ( I108 in IscU ) . Interestingly , deletion of the homologous frataxin gene cyaY in E . coli gave milder phenotypes in that organism than in yeast . Slowed growth , iron accumulation , and oxidative stress sensitivity were not observed in the E . coli knockout [27] in contrast to the yeast knockout . A series of species cross-complementation studies was undertaken in order to test the relative frataxin dependence or independence of yeast expressing the entire E . coli IscU protein targeted to mitochondria . For comparison , IscU protein was reverse engineered to place Met at position 108 , as in the eukaryotic Isu1 . The IscU protein ( authentic E . coli protein with Ile ) and IscU-Met ( substituted E . coli protein with Met ) separately were fused to the leader sequence of yeast mitochondrial protein CoxIV . Each of these fusion constructs was transformed into the GAL1-ISU1/Δisu2 strain . The E . coli IscU proteins , with the authentic isoleucine or with the substituted methionine , were able to function in yeast as indicated by complementation of GAL1-ISU1/Δisu2 cells . Each of these complemented strains grew well , indicating that the E . coli IscU or IscU-Met targeted to yeast mitochondria could function as the only Fe-S cluster assembly scaffold in the cell . No difference between iscU or iscU-Met expressing cells was noted in terms of growth ( Fig 5A , rows 1 and 2 ) . Frataxin was deleted in these strains , and a striking growth phenotype was observed . Both the Δyfh1 [iscU] or Δyfh1 [iscU-Met] could be maintained under an argon atmosphere with subtle differences in colony size on agar plates ( Fig 5A upper panel ) . However , following air exposure , Δyfh1 [iscU] continued to grow with a doubling time of 2 . 5 h , whereas Δyfh1 [iscU-Met] progressively slowed until the doubling time reached 8 . 5 h in defined raffinose-based medium ( Fig 5A , compare top panel for argon growth to bottom panel for aerobic growth in rows 3 and 4 ) . This air/oxygen dependent growth inhibition was much more severe for the Δyfh1 [iscU-Met] strain than for the matched Δyfh1 [ISU1] , carrying Δyfh1 and yeast ISU1-Met . Perhaps the hybrid yeast-E . coli Fe-S cluster assembly machinery is particularly oxygen sensitive in the absence of frataxin . One of the functions of frataxin could be to shield the Fe-S cluster assembly machinery from oxygen [10] . Isolated mitochondria were examined for protein expression by immunoblotting . The yeast Isu1 migrated as a single band of about 14 kDa in mitochondria ( Fig 5B , lane 1 ) . The CoxIV fusions with the E . coli proteins reacted strongly with antibody raised against the yeast Isu1 , giving rise to a slower migrating doublet ( Fig 5B , lanes 2–7 ) . The slower migrating band co-migrated with the bacterial expressed and purified IscU at about 17 kDa , and therefore it likely represents the CoxIV signal sequence-cleaved form of the IscU fusion protein . The more rapidly migrating form may be a proteolytic product . The retarded gel mobility of E . coli IscU compared with yeast Isu1 may be explained by its lower pI ( 4 . 7 for the E . coli IscU versus 9 . 3 for the yeast Isu1 ) , which is associated with decreased binding of SDS and slower migration in the gel [40] . We have observed a similarly aberrant slow migration of Yfh1 due to its many acidic residues and failure to bind SDS [41] . In the Δyfh1 [iscU-Met] mitochondria , IscU protein was markedly increased in abundance ( Fig 5B , lanes 6 and 7 ) . By contrast , in the Δyfh1 [IscU] mitochondria , the IscU protein level was comparable to that of the frataxin plus strains ( Fig 5B , lanes 4 and 5 , compare with lanes 2 and 3 ) . The difference may be traced to defective Fe-S cluster assembly in the Δyfh1 [iscU-Met] cells . In cells with impaired Fe-S cluster assembly , Isu1 protein abundance was previously shown to be up-regulated due to increased iron-dependent transcription mediated by the Aft1/2 regulator , and decreased turnover mediated by the Pim1 protease [32] . Furthermore , the E . coli IscU proteins might be poorly recognized by the yeast Pim1 , further slowing turnover and increasing abundance ( Fig 5B , lane 1 versus lanes 2 and 3 ) . Other mitochondrial proteins were also examined . Nfs1 protein levels were comparable in all cases , consistent with the lack of regulatory changes ( Fig 5B ) . Yfh1 was detected in the cells consistent with the predicted genotypes , being present in YFH1 [ISU1] , YFH1 [iscU] , and YFH1 [iscU-Met] but absent in Δyfh1 [iscU] and Δyfh1 [iscU-Met] ( Fig 5B ) . Cytochrome c , an indicator of cellular heme status , was present in the Δyfh1 [iscU] mitochondria expressing the authentic E . coli protein but completely undetectable in Δyfh1 [iscU-Met] mitochondria expressing the substituted form of the E . coli protein ( Fig 5B , compare lanes 4 and 5 versus lanes 6 and 7 ) . Iron homeostasis was also evaluated . Two independent clones of Δyfh1 [iscU-Met] cells were tested because of the genetic instability and changeable phenotypes associated with this genotype . Both clones exhibited strongly increased cellular iron uptake compared with the unsubstituted control clones Δyfh1 [iscU] ( Fig 5C , Cellular iron , bars 6 and 7 ) . Iron accumulated in the post-mitochondrial supernatant and mitochondria of the Δyfh1 [iscU-Met] strains ( Fig 5C , Post-mito supernatant and Mitochondrial iron , bars 6 and 7 ) . The increase in mitochondrial iron levels was about 2–4 fold more than in the matched Δyfh1 [iscU] strains ( Fig 5C , Mitochondrial iron , bars 4 and 5 ) . Iron accumulated in both soluble and insoluble forms as assessed by centrifugation in the presence of Triton X-100 , probably indicating ferric phosphate nanoparticle accumulation [25 , 26] . In all cases , the Δyfh1 [iscU-Met] mitochondria accumulated more insoluble iron than the Δyfh1 [iscU] mitochondria [Fig 5C , Mitochondrial iron distribution , bars 6 and 7 versus bars 4 and 5 ) . In the frataxin-plus strains expressing E . coli proteins , YFH1 [iscU] and YFH1 [iscU-Met] , iron homeostasis was mostly preserved ( Fig 5C , all panels , bars 2 and 3 ) . The most severe loss of iron homeostasis occurred in the Δyfh1 [iscU-Met] strain and correlated with the concurrent severe deficiency of Fe-S cluster proteins . The mechanism by which defective mitochondrial Fe-S cluster assembly perturbs iron homeostasis is still poorly defined . However , it is likely that important roles are played by loss-of-function of Fe-S cluster binding proteins such as Aft1/2 [42] and glutathione reductases [6] . Aconitase activity , measured by the in-gel assay of mitochondria , was present in Δyfh1 [iscU] mitochondria from two independent clones ( Fig 5D , lanes 4 , 5 and lanes 11 , 12 ) and absent in Δyfh1 [iscU-Met] mitochondria from two independent clones ( Fig 5D , lanes 6 , 7 and lanes 13 , 14 ) . Aco1 protein was present in normal amounts in Δyfh1 [iscU] but markedly decreased in Δyfh1 [iscU-Met] mitochondria , consistent with increased turnover of the apoprotein ( Fig 5B ) . By contrast , in frataxin plus mitochondria , aconitase protein and activity were present in all cases . Aconitase activity was greatest in the frataxin plus ISU1 expressing yeast mitochondria ( Fig 5D , lanes 1 and 8 ) , slightly less in the frataxin plus E . coli iscU expressing mitochondria ( Fig 5D , lanes 2 , 3 and lanes 9 , 10 ) , and slightly less again in the frataxin null iscU expressing mitochondria ( Fig 5D , lanes 4 , 5 and lanes 11 , 12 ) . Thus aconitase activity correlated well with other features of these strains , including growth , cytochrome c levels , and iron homeostasis . Isu1/IscU entries in the public database RefSeq ( 6064 in total ) were collected and aligned on the highly conserved 12 amino acid sequence LPPVK LH CSX LA , using the Muscle algorithm [43] . The sequences were then sorted according to the amino acid at position X , where X is Met in the yeast Isu1 and Ile in the Isu1 suppressor mutant ( S2 Table ) . Sorting on this position revealed highly interesting groupings . Firstly , we found that Isu1/IscU sequences with Met were present almost exclusively in eukaryotic species ( 302 of 307 entries , S2 Table ) . The converse was also true i . e . eukaryotic species had Isu1/IscU with Met present in almost all cases . The proteins with Met at position X were found in the most diverse branches of eukaryotes , including Excavata that lack classical mitochondria , Chromalveolata , various yeasts including Zygomycota , Basidiomycota , Ascomycota , land plants , photosynthetic single-celled organisms , various metazoans including worms , fish , flies , mice and humans ( Fig 6 ) . The only significant exceptions to the rule that Isu1/IscU with Met occurs in eukaryotes were several proteobacterial rickettsial species , including Holospora undulata , Neorickettsia risticii , Neorickettsia sennetsu , and Orientia tsutsugamushi . These organisms are intracellular parasites that are the closest living relatives of mitochondria . Interestingly , all these species have retained frataxin in their genomes [1] , suggesting that IscU-Met and frataxin may have been co-inherited with the rest of the Fe-S cluster assembly machinery during the endosymbiotic event that gave rise to mitochondria ( Fig 6 ) [1] . The lists of Isu1/IscU proteins using Cys , Ile , Leu , and Val at position X , i . e . those amino acid substitutions of yeast Isu1 conferring frataxin-bypass activity , included predominantly prokaryotic organisms . The Cys list had only 6 entries , all from bacteria , including several Clostridium species and an uncultured archeon ( S2 Table ) . For the Ile list , almost all of the species were from prokaryotes ( 171 of 178 entries ) . The exceptions , i . e . eukaryotic species on this list , were interesting in that they often possessed more than one Fe-S cluster assembly scaffold gene . For example , the eukaryotic bumble bee , cucumber , armadillo and a single type of Mediterranean fly each carried Isu1 proteins with Ile at position 141 , but the same organisms possessed another gene which used Met , as is typical for almost all eukaryotes . The Val column had almost exclusively prokaryotic or archaeal species ( 341 of 348 entries ) . In this column , the only eukaryotic species included Entamoeba histolytica and Giardia intestinalis , organisms that are known to have acquired Fe-S cluster assembly components from bacteria by gene transfer [44] . A single eukaryotic plant species , wheat or Aegilops tauschii , was found in this column . However , wheat also had another Isu gene using Met , and so it follows the theme of retaining more than one type of Isu , perhaps for adaptive advantages . Along the same lines , Bos mutus , a yak with a high altitude habitat , carried an Isu with Ile but also retained the more typical eukaryotic Isu with Met . The IscU proteins with leucine were found predominantly in prokaryotic species ( 181 of 199 entries ) . The outlier eukaryotic species on this list included selected apicomplexa ( e . g . Plasmodium falciparum , P . vivax , P . yoelli ) , fungi ( e . g . Cryptococcus and Aspergillus species ) and mitosome containing microsporidia ( e . g . Nematocida parisii ) . The diversity of Isu1/IscU proteins is further increased by the expression of different splice forms in some species . Alternatively spliced forms of the scaffold protein genes were found in humans and armadillo . In humans , the X1 splice form inserts an Arg , and isoform X4 inserts a Lys at position X . The Lys containing isoform has been associated with destabilized protein and development of a human disease characterized by exercise intolerance and mitochondrial myopathy [45] . The substitutions of the Met 141 in yeast Isu1 with Lys and Arg were tested , and neither one was able to support scaffold activity or frataxin-bypass activity ( Fig 1B ) . Nonetheless , it is still possible that these splice forms with Lys or Arg amino acid substitutions could serve a special function or afford an adaptive advantage under some special circumstances or in specific tissues . In summary , Isu1/IscU amino acid sequences with Met at position X of the motif LPPVK LH CSX LA were predominantly found in eukaryotes but also occurred in several Rickettsia species . All contained frataxin in their genomes . Isu1/IscU proteins with the amino acids Cys , Ile , Val or Leu at position X were present primarily in prokaryotes , some with and some without frataxin in their genomes ( Fig 6 ) .
The suppressor ISU1-Ile bypassed the requirement for YFH1 when it was introduced on a plasmid . Similarly , the corresponding substitution in a plasmid carrying the paralogous ISU2-Ile conferred bypass activity . The effects were observed with chromosomal ISU1 and ISU2 genes remaining intact , thereby reflecting genetic dominance and suggesting a gain of function . It is therefore interesting to consider more specifically the nature of the gained function . One explanation could be that Nfs1/Isd11 exhibits only basal cysteine desulfurase activity , and that frataxin is needed to act as a positive effector , thereby inducing the optimal activated level of cysteine desulfurase [13 , 29] . The wild-type Isu1 has no stimulatory activity on its own and may even be inhibitory [13 , 33] , but the suppressor Isu1-Ile is able to substitute for frataxin by stimulating the Nfs1/Isd11 cysteine desulfurase [29 , 30] . If both wild-type Isu1 and suppressor Isu1-Ile are present in the cell simultaneously , the stimulatory effect of the suppressor is the dominant effect , providing bypass activity . An implied consequence of this scenario is that the suppressor Isu1-Ile will require an active form of Nfs1 for it to be effective . Thus it makes sense that the hypomorphic allele of NFS1 , nfs1-14 [46] , with decreased cysteine desulfurase activity , would not support bypass . The results provide genetic support for the hypothesis that frataxin and the bypass Isu1 work to produce their effects on Fe-S cluster assembly , at least in part , by boosting the activity of the cysteine desulfurase . The suppressor Isu1 was shown to stimulate persulfide formation on Nfs1 , similar to frataxin [29 , 30] . Subsequently , sulfur for Fe-S cluster synthesis must be transferred to the Isu1 scaffold and assembled with iron to form the Fe-S cluster intermediate . It is generally assumed that this process occurs in a protein complex with other Isu1 molecules present [35 , 36] . Therefore , we wondered if the Nfs1 stimulatory effect of the suppressor Isu1 could promote Fe-S cluster formation in trans on a normal copy of Isu1 or Isu2 . However , at least in a series of genetic experiments , this was not the case . The survey of all the possible M141 amino acid substitutions identified a set of best scaffolds , moderate scaffolds and poor scaffolds , based on complementing activity in the GAL1-ISU1/Δisu2 strain ( Fig 1B ) . The same set of plasmids , scored for frataxin-bypassing capability , identified activity for M141 with Cys , Ile , Leu or Val changes , and these all fell into the top category for scaffold activity . Furthermore , if a second site substitution abolishing scaffold activity ( e . g . changing a critical Cys to Ala ) was introduced into the M141I-Isu1 protein sequence , and the doubly substituted Isu1 was introduced into a Δyfh1 shuffle strain with wild-type ISU1 and ISU2 present , bypass activity was abrogated ( Fig 1C ) . Thus most likely the suppressor Isu1 does not productively interact with the wild-type Isu1 to mediate bypass , but instead it replaces the wild-type Isu1 in providing both bypass and scaffold functions . Eukaryotic and prokaryotic Fe-S cluster machineries are highly conserved . Both yeast and E . coli utilize cysteine desulfurases , scaffold proteins and frataxin homologs . However major differences in the frataxin deletion phenotypes have been reported , with essentiality or severely deleterious phenotypes in the eukaryotic case [4] , and normal growth and relatively mild phenotypes in E . coli [27] . Why the difference ? One possibility is that E . coli possesses a redundant Fe-S cluster assembly system , the SUF system , which may compensate for lack of frataxin [47] . However this does not entirely account for the phenotypic differences , because the SUF system is not generally deployed under standard growth conditions . Significantly , the cysteine desulfurases show key differences . Most prokaryotic cysteine desulfurases such as IscS are constitutively active , although some regulatory changes in activity have been described [15] . The eukaryotic cysteine desulfurases , on the other hand , seem to be largely inactive in their basal state . Activation is required , and this activation involves frataxin . Purified Nfs1 is able to bind the substrate cysteine in its PLP containing substrate-binding site . However , binding is inefficient , and frataxin interaction increases exposure and utilization of substrate-binding sites of the enzyme [29] . The suppressor Isu1-Ile protein is able to generate a similar alteration of Nfs1 , mimicking the effect of frataxin on Nfs1 , and providing a plausible explanation for its frataxin-bypass activity . Nfs1 enzyme with the substrate bound must still undergo another activation step , mediated by Isd11 . Isd11 triggers a conformational change and persulfide formation , thereby generating the intermediate for Fe-S cluster assembly [29] . Here we have seen that not only the Ile but also the Cys , Val , Leu substituted forms of Isu1 are able to stimulate persulfide formation on Nfs1 in the absence of frataxin . Thus these bypassing alleles of Isu1 activate Nfs1 independently of frataxin , rendering the mitochondria more prokaryote-like . More than 17 , 000 amino acid changes of ISU1 were surveyed , but only a small subset of those conferred bypass activity . In all cases , the active changes altered the amino acid at position 141 of Isu1 , introducing the amino acids Cys , Ile , Leu , or Val . The mutagenesis was not exhaustive , but nonetheless , the data support the very restricted nature of the changes that confer this activity . The biochemical features of these newly discovered bypass mutants ( Isu1 with Cys , Leu , or Val substituted at position 141 ) were similar to the original one ( Isu1 with Met replaced by Ile ) . The various mutant forms of Isu1 conferred improved growth in the absence of frataxin and more efficient Fe-S cluster assembly in mitochondria . The persulfide-forming activity was increased in mitochondria lacking frataxin , indicating enhanced cysteine desulfurase activity . How does the Isu1 suppressor work ? Isu1 is a central component of the Fe-S cluster assembly complex consisting of Nfs1/Isd11/Isu1/Yfh1 . The components are highly conserved with their bacterial homologs , with the exception of Isd11 . Structural information has been obtained only for the bacterial components [11 , 36] . For Isu1/IscU the structure includes alpha helices framing a platform of beta sheets , with three conserved cysteines oriented towards a binding pocket in the core and able to coordinate the Fe-S cluster intermediate ( Fig 7 ) . The amino acid motif LPPVK is found towards the beginning of a long C-terminal alpha helix [11 , 39] . Interestingly , the PVK motif ( Fig 7 , green in Fe-S scaffold ) was shown to include the frataxin-binding site [33] . The suppressor residue Ile ( Fig 7 , red ball-and-stick ) is predicted to lie on an exposed surface of this helix , opposite the Fe-S liganding Cys , which is found on the opposite interior face of the helix ( Fig 7 , blue ball-and-stick ) . Thus the Met to Ile amino acid change near to this frataxin-binding site on Isu1 might mimic the effects of frataxin binding . Substitutions of Cys , Val , or Leu would be predicted to produce similar changes . Conformational changes of the Fe-S cluster assembly complex facilitating Fe-S cluster intermediate formation might ensue . Nfs1 might be altered in a way that exposes substrate-binding sites for interacting with cysteine [29] . ( Fig 7 , yellow balls for PLP in the binding site ) . The sulfur intermediate , converted to a persulfide and bound to the flexible loop of Nfs1 ( Fig 7 , magenta dotted line and adjacent residues ) , might be more readily transferred to the modified Isu1 . The Cys 139 , an Fe-S cluster ligand on Isu1 , that is one helix turn away from the Ile suppressor residue , might be rendered more accessible for persulfide transfer ( Fig 7 ) . Iron delivery remains the least characterized step in Fe-S cluster formation . Frataxin might facilitate iron entry into the Fe-S cluster assembly complex by initiating a conformational change that promotes transfer of mitochondrial iron from a physiological ligand ( still undefined ) to Isu1 . Alternatively , frataxin might play a more direct role in binding iron and delivering it to Isu1 as has been shown for Yfh1 [19] . The Isu1-Ile protein ( or bacterial IscU ) might accomplish a similar function by exposing iron-binding sites on Isu1 [19] . The iron delivery step in Fe-S cluster assembly will need to be better understood to support or to refute these possibilities . Iron homeostasis was improved , correlating with the improvements in cytochrome c , a mitochondrial heme protein . In previously published work , cytochromes were virtually absent in Δyfh1 and restored by the ISU1-Ile allele as assessed by low temperature spectra of whole cells [28] . Other heme proteins such as cytochrome c peroxidase , Ccp1 , were similarly affected [30] . Heme synthesis , measured by 55Fe incorporation into porphyrin in isolated mitochondria , was very low in Δyfh1 and recovered in the presence of the substituted ISU1 [28] . What is the mechanistic connection between a change in the Fe-S cluster assembly scaffold and these effects on heme synthesis ? Heme synthesis occurs inside mitochondria and involves the insertion of iron into porphyrin by the enzyme ferrochelatase to make protoheme . Porphyrin and ferrochelatase activity are not lacking , and in fact , zinc protoporphyrin accumulates in the Δyfh1 mutant [25] . Instead iron is the likely source of the trouble . Iron accumulating in Δyfh1 mitochondria ( and in other Fe-S cluster deficient cells ) exhibits changes in its physical properties and solubility that could lead to loss of bioavailability [25] . The data suggest that Fe-S cluster synthesis acts upstream of heme synthesis in promoting iron bioavailability for heme synthesis , although many mechanistic details are still lacking . Isu1/IscU amino acid sequences sorted according to the amino acid appearing in position 141 fall into clearly defined grouping , underscoring the importance of this residue for Isu1/IscU function . The amino acid Met appeared almost exclusively in eukaryotic organisms . The only significant exceptions were several prokaryotic Rickettsia species , which were classified as proteobacteria . In all of these organisms with IscU-Met , frataxin was also present in their genomes . We can imagine a scenario in which IscU-Met and frataxin originated together in proteobacterial ancestors of mitochondria such as Rickettsia , and from there gave rise to modern mitochondria via horizontal gene transfer during the endosymbiotic event ( Fig 6 ) . A key point is that the Met amino acid was found only in organisms with frataxin , as if Met serves to “lock in” frataxin by ensuring frataxin dependence of Fe-S cluster assembly . The co-dependence of these two elements , the IscU-Met and frataxin , appears to be quite profound , as they have been retained together throughout all the eukaryotic branches of the tree of life . In the prokaryotic world , on the other hand , various IscU variants were found , and various amino acids were found at position X adjacent to the PVK frataxin-binding motif of the IscU homologs . The amino acids Cys , Ile , Leu , or Val which conferred frataxin-bypass in biochemical experiments in yeast , were found in IscU proteins of species with or without frataxin homologs ( Fig 6 ) . Thus the evolutionary record suggests that these amino acids may be associated with relative frataxin independence of Fe-S cluster assembly . The advantages of retaining IscU-Met and frataxin versus IscU-Ile and no frataxin remain to be ascertained , as both arrangements are able to support efficient and regulated Fe-S cluster assembly . Friedreich’s ataxia is a progressive degenerative disease affecting neurons such as dorsal root ganglia and cardiomyocytes , and certain other tissues . The disease is caused by frataxin deficiency in affected tissues and is associated with defective Fe-S cluster assembly [24 , 48] . The efficacy of frataxin-bypass in yeast can be viewed as a reprogramming of mitochondrial Fe-S cluster assembly to a more prokaryotic type , such that the cysteine desulfurase and other features of the assembly process become more frataxin-independent . The discovery of several substitutions , changes to Cys , Ile , Leu , or Val , that are able to confer frataxin-bypass suggests that there may be common structural features of these Isu proteins that could be mimicked by small molecules [49] . The genetic dominance of the suppressor activity in the Δyfh1 yeast also suggests that such an approach could be effective in a therapeutic setting in human mitochondria where frataxin is deficient and normal Isu1 is still present .
A set of plasmids was generated containing modified versions of the Isu1 coding sequence ( between NdeI and XhoI ) , carried between the native Isu1 promoter ( 700 bp between EagI and NdeI ) and terminator ( 200 bp between BamHI and SacI ) on a centromere based plasmid , YCplac22 . Residue 141 of the full length Isu1 precursor protein was changed from M to each of the other 19 amino acids , using QuikChange mutagenesis ( Agilent Technologies Inc . ) . In addition , N123D and N123A mutants were generated . Cysteine mutants of the Isu1 coding sequence were constructed in which C69 , C96 and C139 were each changed to A . The M141I change was then introduced into each of these cysteine mutants , creating plasmids C69A-M141I , C96A-M141I , and C139A-M141I . A cysteine swap mutant was created in which C139A was combined with M141C in the full length Isu1 . The various mutant forms of Isu1 were tested for Yfh1-bypassing function and scaffold function by transforming into strains YFH1 shuffle and GAL1-ISU1/Δisu2 , respectively ( Table 1 ) . For testing genetic dominance of the ISU1 bypass suppressor , the suppressor allele ( YCplac22-ISU1-M141I ) was introduced into strain 70–31 containing genomic copies of ISU1 and ISU2 ( Table 1 ) , and the covering YFH1 plasmid was removed by fluoroorotic acid ( FOA ) treatment . For testing ISU2 function , the native ISU2 sequence was amplified from plasmid pGP564-ISU2 including 700 bp 5’ and 200 bp 3’ of the coding sequence , and cloned between restriction sites HindIII and BamHI in YCplac22 , making plasmid YCplac22-ISU2 . The M133 was changed to Ile by site directed mutagenesis , creating YCplac22-ISU2-M133I . The ISU2 coding sequence was inserted into NdeI-XhoI sites in place of the ISU1 coding creating YCplac22-ISU2coding . The M133 was changed to Ile by site directed mutagenesis creating plasmid YCplac22-ISU2coding-M133I . Strain GAL1-ISU1/Δisu2 was transformed with plasmid YCplac22-ISU1 in which M141 was changed to C , I , L or V and the chromosomal GAL1 promoter was turned off by shifting cells from galactose to glucose as the carbon source . The YFH1 gene was deleted by transforming with plasmid pRS405-gamma-yfh1 linearized with BamHI , and selecting for transformants on defined leucine drop-out medium in an argon-filled chamber . These low oxygen conditions were previously shown to mitigate the mutant phenotype and allow for stable propagation of the knockouts [30] . The knockouts were confirmed by PCR of the YFH1 locus . The strains were denoted Δyfh1 [ISU1] ( 115–26 ) , Δyfh1 [ISU1-Cys] ( 116–54 ) , Δyfh1 [ISU1-Ile] ( 115–28 ) , Δyfh1 [ISU1-Leu] ( 116–53 ) , and Δyfh1 [ISU1-Val] ( 116–51 ) . Congenic wild-type strain YFH1 [ISU1] was included as a control . The strains were thawed from -80°C vials and inoculated to CSM-Trp/2% raffinose defined medium agar plates kept in an argon-filled jar . Cells were inoculated from the plates into liquid medium of the same composition . In general , small cultures of 50 ml were grown in argon-filled bottles for two days without shaking , expanded to 100 ml cultures while shaking in air , and then diluted again into 1 L cultures supplemented with 10 μM ferrous ascorbate . The 1 L cultures were grown at 30°C for 16 h , and the total time exposed to air was approximately 24 h for the slow-growing strains . Cellular and subcellular iron levels were determined by growing cells for at least four doublings in standard defined medium supplemented with 58 nM 55FeCl3 , 10 μM unlabled ferric chloride and 100 μM ascorbic acid . Cells were washed free of unincorporated iron , and then they were ruptured by vortexing with glass beads in the presence of 50 mM Hepes/KOH , pH 7 . 5 , 150 mM NaCl , 0 . 6 M sorbitol . Differential centrifugation was used to remove unbroken cells and to separate the remainder into mitochondrial and post-mitochondrial fractions [50] . The post-mitochondrial fraction included both cytoplasm and vacuoles , as no effort was made to separate these two cellular components . Mitochondria were shown to be mostly intact by evaluation of mitochondrial marker proteins , which remained with the mitochondrial fraction . Mitochondria were lysed for 10 min at room temperature in the presence of 0 . 1% Triton X-100 in hypotonic buffer ( 50 mM Hepes/KOH , pH 7 . 5 , 150 mM NaCl ) . The supernatant ( soluble ) and pellet ( insoluble ) portions were separated by centrifugation at 20 , 000 x g for 30 min . Iron content was determined by scintillation counting for 55Fe , and protein content was measured by bicinchoninic acid assay ( BCA , Pierce ) [28] . For whole cells , iron content was reported as pmol iron per million cells , whereas for the cellular fractions iron content was reported as pmol iron per microgram protein . An in-gel activity assay for aconitase was performed . Briefly , mitochondria were lysed in buffer consisting of 50 mM Tris-HCl pH 8 , 50 mM NaCl , 1% TX-100 , 10% v/v glycerol , 2 mM Na-citrate and 15 U catalase . Samples were loaded on a native acrylamide gel containing 132 mM Tris base , 132 mM boric acid , 3 . 6 mM sodium citrate . The signal was developed in the gel by incubating in developing buffer containing 100 mM Tris-HCl , pH 8 , 1 mM NADP , 2 . 5 mM cis-aconitic acid , 5 mM MgCl2 , 1 . 5 mM methylthiazolyldiphenyl-tetrazolium bromide ( MTT ) , 0 . 3 mM phenazine methosulfate , and 5 U/ml isocitrate dehydrogenase [51 , 52] . Persulfide formation on Nfs1 present in intact mitochondria was measured as described [16] . Isolated mitochondria were depleted for endogenous nucleotides and NADH by incubation for 10 min at 30°C in buffer ( 20 mM Hepes/KOH , pH 7 . 5 , 0 . 6 M sorbitol ) . These mitochondria were incubated with 35S-cysteine ( 10 μCi ) for 15 min at 30°C . Mitochondria were recovered , proteins were separated by non-reducing SDS gel , and the persulfide was viewed by radioautography . Fe-S cluster formation was measured as described [12] . Mitochondria were incubated with 35S-cysteine for 30 min in the presence of added ATP ( 4 mM ) , GTP ( 1 mM ) , NADH ( 5 mM ) and iron ( 10 μM ferrous ascorbate ) . Mitochondria were recovered and the soluble proteins were released by freeze-thaw and sonication and separated on a native gel . The signal associated with newly formed [Fe-35S] aconitase was visualized by radioautography . Pools of mutagenized linear fragments of 1443 bp including the coding region of the ISU1 gene were generated by mutagenic PCR in the presence of 1 mM MnCl2 and altered nucleotide concentrations ( 2 mM dATP , 2 mM dGTP , 10 mM dTTP , 10 mM dCTP ) . The primers used were: A ( 5’ TTTTTTCGGCCGTTCTTTTCTTTTTCTTGCACTACC 3’ ) and B ( 5’ TGATTTGAGCTCagcacgtccgtcccgctttcaccctgg 3’ ) , and the templates were different YCplac22-ISU1 plasmids with M , Y , H or F at position 141 of the coding region . Each mutagenized pool was co-transformed with a gapped plasmid ( pRS416-ISU1 digested with MscI and BamHI ) into the YFH1 shuffle strain 109–9 ( Δyfh1::TRP1 Δisu1::HIS3MX6 ISU2 cyh2 [pRS318-CYH2-LEU2-YFH1] , Table 1 ) . After selecting for transformants on uracil drop-out medium with glucose as the carbon source , the colonies were replicated to uracil drop-out , cycloheximide medium with glucose or raffinose as the carbon source . The large colonies appearing after several days on the raffinose plates were analyzed further . Colony PCR was used to check for absence of YFH1 . Transformants still harboring YFH1 after counterselection were discarded . The remaining clones were expanded , and the plasmid-borne ISU1 alleles were rescued in E . coli and the DNA was sequenced . The coding region of E . coli IscU was amplified from genomic E . coli DNA from strain DH5 alpha and inserted into the XbaI and XhoI sites of a YCplac22 derived plasmid . In this plasmid , the ISU1 promoter consisting of 700 bp , is followed by the first 22 amino acids of the cytochrome c oxidase subunit IV ( CoxIV ) mitochondrial signal sequence [53] , and 200 bp of ISU1 terminator . The resulting plasmid was checked by DNA sequencing and confirmed to code for a CoxIV-IscU fusion protein with amino terminus as follows: MLSLRQSIRFFKPATRT^LCSSRHMAYSEKVID . The caret indicates the predicted signal sequence cleavage site by the mitochondrial processing peptidase , and the bolded letters indicate amino acids of E . coli IscU . The rest of the IscU sequence was confirmed to be the same as listed in Genbank accession No . AAJU02000016 . 1 . The amino acid I108 of IscU ( equivalent to M107 in signal-cleaved mature Isu1 or M141 in the full length precursor form of Isu1 ) was changed from I to M by QuikChange mutagenesis , generating a plasmid for expressing IscU-Met in mitochondria . The mitochondrial-targeted iscU plasmids were introduced into strain GAL1-ISU1/Δisu2 , generating strains YFH1 [iscU] and YFH1 [iscU-Met] . YFH1 was deleted in these strains by transforming with pRS405-gamma-yfh1 , linearized at BamHI , followed by selection on leucine drop-out medium in an argon-filled anaerobic jar . This created strains Δyfh1 [iscU] and Δyfh1 [iscU-Met] ( Table 1 ) . Deletion of YFH1 was verified by PCR . Strain GAL1-ISU1/Δisu2 transformed with YCplac22-ISU1 served as the control strain YFH1 [ISU1] . All strains were grown in an argon-filled chamber or in argon-bubbled medium , and cells were then exposed to air in defined raffinose medium for iron labeling and mitochondrial isolation . Isu1/IscU related sequences were collected by using BLAST similarity to the Isu1 of Saccharomyces cerevisiae S288c from the RefSeq non-redundant database . All related entries in RefSeq ( 6064 ) were aligned using the Muscle program [43] , according to the amino acid at position X ( equivalent to Isu1 residue 141 ) of the amino acid motif LPPVK LH CSX LA . The lists were manually edited , and duplicates were removed , retaining one entry for each species . For each entry , the sequence was scored + if it contained 8 or more amino acids identical to the query , and 0 if contained 7 or less . The sequences were scored * for exceptions if they deviated from the rule that eukaryotic Isu proteins use only methionine at that amino acid position and prokaryotic Isu proteins do not use methionine at that position . | Frataxin was discovered because mutations in the corresponding gene cause the neurodegenerative disease Friedreich’s ataxia . The finding that frataxin protein physically associates with scaffold proteins Isu1/IscU places it squarely in the pathway of Fe-S cluster assembly . Fe-S clusters are essential cofactors for many proteins involved in cellular respiration , DNA repair , translation and other processes . Frataxin is conserved throughout evolution , being present in eukaryotes such as yeast and human and in some prokaryotes including E . coli . However , differences exist between the eukaryotic and prokaryotic forms of frataxin . The eukaryotic forms are critical for Fe-S cluster assembly whereas prokaryotic forms are more dispensable . We found that a key to this difference is a single amino acid in the scaffold protein Isu1 at position 141 . Changes of the eukaryotic amino acid , Met , to prokaryotic amino acids , Ile , Leu , Cys , or Val , rendered mitochondria more frataxin-independent . No other changes were able to replicate this effect . Thus , Isu1 containing Met at position 141 may have coevolved with frataxin in eukaryotes , conferring frataxin-dependence . In contrast , the appearance of other amino acids at this position may have rendered prokaryotic cells less dependent on frataxin . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Turning Saccharomyces cerevisiae into a Frataxin-Independent Organism |
Gamma-herpesviruses persist in lymphocytes and cause disease by driving their proliferation . Lymphocyte infection is therefore a key pathogenetic event . Murid Herpesvirus-4 ( MuHV-4 ) is a rhadinovirus that like the related Kaposi's Sarcoma-associated Herpesvirus persists in B cells in vivo yet infects them poorly in vitro . Here we used MuHV-4 to understand how virion tropism sets the path to lymphocyte colonization . Virions that were highly infectious in vivo showed a severe post-binding block to B cell infection . Host entry was accordingly an epithelial infection and B cell infection a secondary event . Macrophage infection by cell-free virions was also poor , but improved markedly when virion binding improved or when macrophages were co-cultured with infected fibroblasts . Under the same conditions B cell infection remained poor; it improved only when virions came from macrophages . This reflected better cell penetration and correlated with antigenic changes in the virion fusion complex . Macrophages were seen to contact acutely infected epithelial cells , and cre/lox-based virus tagging showed that almost all the virus recovered from lymphoid tissue had passed through lysM+ and CD11c+ myeloid cells . Thus MuHV-4 reached B cells in 3 distinct stages: incoming virions infected epithelial cells; infection then passed to myeloid cells; glycoprotein changes then allowed B cell infection . These data identify new complexity in rhadinovirus infection and potentially also new vulnerability to intervention .
Herpesviruses are among the most prevalent of all persistent pathogens . Thus even when disease is individually rare , the total burden in populations is large . The difficulty of eliminating latent viral genomes makes latency establishment an important target for infection control . Gamma-herpesviruses persist in lymphocytes . Epstein-Barr virus ( EBV ) infects B cells better than epithelial cells in vitro , and prominently colonizes tonsillar B cells during acute infectious mononucleosis [1] . Thus EBV [2] and the Kaposi's Sarcoma-associated Herpesvirus ( KSHV ) [3] have been proposed to infect tonsillar B cells directly after oral host entry . However infectious mononucleosis post-dates EBV host entry by at least a month [4] . Therefore the infection seen at that time may correspond to host exit rather than entry , and vaccination to prevent B cell infection failed to reduce EBV seroconversion rates [5] . One barrier to understanding gamma-herpesviruses solely through EBV and KSHV is that their narrow species tropisms limit in vivo analysis . Experimentally accessible gamma-herpesviruses such as Murid Herpesvirus-4 ( MuHV-4 ) [6]–[9] consequently provide an important source of information . MuHV-4 is closely related to KSHV [10] , [11] . Like EBV and KSHV it persists in B cells [12] . It also infects myeloid cells [13] . Most experimental infections have delivered MuHV-4 intranasally to mice under general anesthesia; aspirated virions then infect lung epithelial cells [14] . The detection by PCR of replication-deficient viral DNA from flow cytometrically sorted lung B cells in this setting led to the idea that B cells are a direct infection target [15] , [16] . However viral DNA+ B cells were not detected in lymphoid organs , and adsorbed inoculum debris was not excluded as the source of viral DNA . A further caveat to drawing general conclusions from lung infection is that MuHV-4 inhaled without anesthesia does not reach this site: it replicates just in the nose [17] before following a CD11c-dependent route to lymphoid tissue [18] . Our understanding of human herpesvirus infections is based largely on analysis in vitro . A key task with MuHV-4 is therefore to relate in vitro tropisms to host colonization . Fibroblast-propagated MuHV-4 efficiently infects mice [17] , [19] yet like KSHV seems to infect B cells poorly: despite reports of MuHV-4 infected B cell lines [20] , [21] and phenotypic changes in virus-exposed B cells [22] , [23] , efficient B cell infection has not been demonstrated . MuHV-4 depends on heparan sulfate ( HS ) to infect adherent cells [24] , and poor B cell infection by MuHV-4 and KSHV has been attributed to B cells lacking HS [25] . However infection was not convincingly demonstrated even when B cell HS expression increased . Therefore the barriers to B cell infection remain ill-defined . Here we found no evidence for direct mucosal B cell infection by MuHV-4 entering the upper respiratory tract . Host entry was instead an epithelial infection . This corresponded to in vitro B cell infection showing binding and post-binding blocks . Unlike B cell infection , myeloid infection was limited only by binding and worked well by co-culture with infected fibroblasts . B cell infection improved only when virions came from myeloid cells . These virions showed a constitutive triggering of entry-associated changes in gB and gH . Myeloid cells were closely associated with the acutely infected epithelium , and cre/lox virus marking showed that most of the virus reaching lymphoid tissue had passed through cells expressing CD11c and myeloid-specific lysozyme ( lysM ) . Thus we propose that rhadinoviruses entering new hosts infect epithelial cells , then myeloid cells , and only then B cells .
MuHV-4 is non-infectious orally , but readily infects via the upper respiratory tract [17] . The murine nasal-associated lymphoid tissue ( NALT ) is analogous to human tonsils [26] and provides a potential target for lymphotropic viruses entering the nasopharynx . To determine whether MuHV-4 targets the NALT , we allowed unanesthetized mice to inhale virus and visualized infection 6 days later by immunostaining with a polyclonal , MuHV-4-specific rabbit serum ( Fig . 1a–c ) . Lytic antigens were abundant in the olfactory neuroepithelium ( Fig . 1a ) but absent from the NALT and its overlying epithelium ( Fig . 1b ) . Most of the neuroepithelium is anterior to the NALT , and so potentially more accessible , but even when neuroepithelial infection was evident in the same histological section ( Fig . 1c ) the NALT lacked viral antigens . Our immune serum predominantly recognizes MuHV-4 lytic antigens [27] , so it remained possible that the NALT was latently infected . We tested this by in situ hybridization for the viral tRNA/miRNAs that are abundantly expressed in latently infected splenic B cells [28] . Again neuroepithelial infection was readily identifiable ( Fig . 1d ) but NALT infection was not ( Fig . 1e ) , even when neuroepithelial cells in the same histological section were tRNA/miRNA+ ( Fig . 1f ) . At 14 days post-infection viral tRNA/miRNA expression was abundant in the NALT ( Fig . 1g ) . However this post-dates the virus spread and amplification associated with infectious mononucleosis [7] , and infection was accordingly abundant also in lymph nodes and the spleen ( Fig . 1g ) . Thus the primary i . n . infection was epithelial , and NALT infection did not occur until there was systemic virus spread . To understand why the NALT was not acutely infected we tested virions for their capacity to infect B cells . A key point was to reliably identify B cell infection . Viral tRNA/miRNA detection is not easily combined with staining for cell type-specific markers , and while some infected B cells express ORF73 and M2 [29] , little ORF73 seems to be made [30] and M2 expression is unlikely to be universal . We therefore used an intergenic EF1α promoter to express constitutively viral eGFP ( Fig . S1 ) . Mice infected with EF1α-eGFP+ MuHV-4 showed eGFP expression in 1 . 7% of CD19+ lymph node B cells , consistent with PCR-based estimates of wild-type virus loads [29] , [31] , [32]; this virus also labelled more convincingly than did HCMV IE1-eGFP+ MuHV-4 A20 B cells over-expressing the HS carrier syndecan-1 . It therefore provided a good basis for detecting B cell infection . We next used EF1α-eGFP+ MuHV-4 to compare the infectibility of different cell types . For consistency with studies using other read-outs [20] , [21] , [24] , [33] we tested CHO-K1 epithelial cells , RAW-264 myeloid cells , and NS0 and A20 B cells , exposing each to cell-free virions and assaying infection 18 h later by flow cytometry ( Fig . 2 ) . A particular question was how far cellular HS expression limits infection . MuHV-4 HS dependence is due in part to the inhibitory effect of its gp150 , as gp150 null mutants are much less HS-dependent in binding and infection than the wild-type [24] . We therefore compared infection by gp150+EF1α-eGFP+ and gp150−EF1α-eGFP+ virions . The gp150− and gp150+ virions infected HS+ CHO epithelial cells similarly . The gp150− virions infected HS− CHO cells much better - approximately 30-fold fewer virions than wild-type gave an equivalent number of eGFP+ cells ( Fig . 2a , 2b ) . They also infected RAW-264 monocytes better ( Fig . 2c ) , arguing that poor HS expression limits myeloid infection . In contrast , both gp150− and gp150+ virions infected A20 B cells poorly ( <1% eGFP+ at 10 p . f . u . /cell ) . The B cell lines WEHI-231 and BCL-1 also showed <1% eGFP expression after exposure ( 10 p . f . u . /cell , 18 h ) to gp150+EF1α-eGFP+ or gp150−EF1α-eGFP+ virions ( data not shown ) . Therefore poor B cell infection could not be explained simply by a lack of HS . NS0 myeloma cells were infected better than A20 , WEHI-231 or BCL-1 , particularly by gp150− virions , but remained less infectible than RAW-264 or HS− CHO cells . Therefore all B cell-derived lines showed an infection block beyond poor HS expression . We used flow cytometry to quantitate cellular HS display ( Fig . 3 ) . MAb F58-10E4 recognizes a sulfation-dependent epitope [34] , while mAb NAH46 recognizes a sulfation-independent epitope [35] . Both recognize most forms of HS because it typically shows partial sulfation [36] . Neither mAb stained HS− CHO cells , nor showed more than minimal staining of RAW-264 and A20 cells . NAH46 strongly stained HS+ CHO cells and NS0 cells , whereas F58-10E4 stained HS+ CHO cells strongly and NS0 cells only weakly . Therefore CHO cells expressed partially sulfated HS; NS0 cells expressed largely unsulfated heparan; and A20 and RAW-264 cells expressed little of either form . What HS forms MuHV-4 binds to are unknown . We therefore measured functionally relevant HS display by staining cells with an Fc fusion of the viral gp70 HS binding domains ( gp70-SCR1-3-Fc ) [33] ( Fig . 3 ) . An Fc fusion of the MuHV-4 gp150 residues 1–250 , which does not detectably bind to cells [37] , provided a negative control . Gp70-SCR1-3-Fc binding was more sensitive than mAb binding , possibly because it binds to a wider range of HS modifications , but like mAb NAH46 it bound HS+ CHO and NS0 cells well , RAW-264 and A20 cells less well , and HS− CHO cells hardly at all . Thus NS0 cells were infected poorly despite displaying HS for virus binding . We next tested cell binding by virions , using gp150+ and gp150− versions of MuHV-4 made fluorescent by an eGFP tag on the abundant envelope component gM [38] ( Fig . 4 ) . Gp150−gM-eGFP+ virions bound better than gp150+gM-eGFP+ to all cell types , but the difference was most marked for HS− CHO cells , A20 B cells and RAW-264 monocytes . Thus gp150+ virion binding correlated with cellular HS display , and gp150− virion binding was strong regardless . Both gp150− and gp150+ virions bound well to NS0 cells . Comparing Fig . 2 with Fig . 4 shows that MuHV-4 infected RAW-264 and HS− CHO cells better than NS0 despite binding better to NS0 cells , and that gp150− virions infected A20 cells poorly despite binding relatively well . These data therefore supported the idea of a post-binding block to B cell infection . That gp150− virions bound better than gp150+ to HS+ CHO cells ( Fig . 4 ) but did not infect them better ( Fig . 2 ) suggested that when cell binding was very strong , down-stream events could also limit epithelial infection . However they limited A20 cell infection even when HS expression ( Fig . 3 ) and virion binding ( Fig . 4 ) were weak . By comparison , HS− CHO cells and RAW-264 cells showed a good correlation between better gp150− virion binding and better infection . Therefore a post-binding restriction of infection was possible for any cell , but was much more severe for B cells . We tested further the relationship between cellular HS expression and MuHV-4 infection by expressing in RAW-264 and A20 cells an uncleavable form of the HS carrier syndecan-1 ( SDC-1 ) ( Fig . 5 ) . This increased gp150− as well as gp150+ virion binding , presumably because gp70 and gH/gL attach virions better than does just the HS-independent binding regulated by gp150 . Gp70-SCR1-3-Fc bound only marginally better to RAW-264-SDC-1 cells than to RAW-264 ( Fig . 5a ) , but virion binding ( Fig . 5b ) and infection ( Fig . 5c ) both increased . By contrast , both gp70-SCR1-3-Fc ( Fig . 5a ) and virions ( Fig . 5b ) bound substantially better to A20-SDC-1 cells than to A20 , but infection remained negligible ( Fig . 5c ) . Therefore again there was evidence of a post-binding block to B cell infection . Probably most in vivo herpesvirus spread occurs through cell/cell contacts rather than cell-free virion release [39] . Consistent with this idea , MuHV-4 lacking gp150 accordingly spreads normally in vivo despite poor virion release [27] , whereas MuHV-4 lacking gp48 , which is impaired in cell/cell spread , is attenuated [40] . Thus as B cells were infected down-stream of host entry ( Fig . 1 ) , we reasoned that their infection might involve cell/cell contact . However A20 B cells co-cultured overnight 1∶1 with infected BHK-21 cells ( 1 p . f . u . /cell gp150+EF1α-eGFP+ or gp150−EF1α-eGFP+ virus , 24 h ) remained <1% eGFP+ ( 0 . 10±0 . 02% eGFP+ for gp150+ and 0 . 31±0 . 04% eGFP+ for gp150− ) ( mean ± SD of triplicate cultures ) . By contrast RAW-264 monocytes co-cultured with infected BHK-21 cells became 40–60% eGFP+ ( Fig . 6a ) . RAW-264 cells were infected approximately 30-fold better by cell-free gp150− virions than by gp150+ ( Fig . 2a ) ; co-cultures showed only a 3-fold difference , presumably because cell/cell contact made virion binding less HS-dependent . Therefore direct contact with infected fibroblasts allowed efficient myeloid infection . The contrasting failure to improve B cell infection was consistent with this having an additional , post-binding block that required a different solution . That MuHV-4 targets CD11c+ cells in lymph nodes [18] suggested that myeloid infection might make MuHV-4 B cell-tropic . We therefore next co-cultured A20 B cells 1∶1 with MuHV-4-exposed RAW-264 monocytes ( 3 p . f . u . /cell , 18 h ) . In contrast to the negligible effect of co-culture with infected BHK-21 cells , this led to 6 . 15±1 . 28% of A20 cells becoming eGFP+ for gp150+ MuHV-4 and 12 . 79±2 . 23% becoming eGFP+ for gp150− ( mean ± SD of triplicate cultures ) . Co-culture with RAW-264 cells therefore promoted A20 B cell infection . Similar results were obtained with splenic B cells ( Fig . 6b ) . We could also infect splenic B cells by co-culture with MuHV-4-infected peritoneal macrophages ( Fig . 6c ) . As a further measure of infection we used a MuHV-4 derivative ( MuHV-RG ) in which cre recombinase switches reporter gene expression: MuHV-RG is mCherry+ until it infects cre+ cells , when it becomes eGFP+ ( Fig . 6d , 6e ) . Reporter gene expression is from a lytic cycle promoter , so as NS0 B cells support viral lytic gene expression [13] we co-cultured cre+ NS0 cells with MuHV-RG-exposed cre− BHK-21 cells or cre− RAW-264 cells . Co-culture with RAW-264 cells gave more eGFP+ NS0 cells than did co-culture with BHK-21 cells ( Fig . 6f ) . Therefore myeloid infection promoted B cell infection . Co-culture infections are complicated and so difficult to dissect further . We therefore tested next whether cell-free virions derived from myeloid cells could also infect B cells ( Fig . 7 ) . Gp150+ virions from RAW-264 cells gave greater eGFP expression in splenic B cells than did those from BHK-21 cells; RAW-264-derived gp150− virions worked substantially better ( Fig . 7a , 7b ) , indicating that without cell/cell contact gp150 significantly inhibited B cell binding . Viral DNA quantitation by Q-PCR ( Fig . 7c ) showed more gp150− than gp150+ binding to B cells , with little difference between RAW-264 cell-derived and BHK-21 cell-derived virions . Infectious centre assays and eGFP expression by contrast showed considerably more infection by RAW-264 cell-derived virions . Therefore virion passage through myeloid cells improved B cell penetration rather than B cell binding . Immunoblotting showed differences in gB and gp150 between BHK-21 cell-derived and RAW-264 cell-derived virions ( Fig . 8a ) . However these seemed unlikely to account for their different tropisms: gB N-terminus recognition by mAb MG-2C10 , which depends on cell type-specific O-glycosylation , was reduced for RAW-264 cell-derived virions but deleting the gB N-terminus has no obvious effect on B cell colonization [41]; gp150 migration and recognition , which also vary with O-glycosylation [42] , were different between BHK-21 and RAW-264 cell-derived virions but gp150 disruption enhanced B cell binding by both without being sufficient for infection ( Fig . 7 ) . An important feature of virions not revealed by immunoblotting is that gB and gH - which drive herpesvirus membrane fusion - change in antigenicity during cell entry [43] , [44] . The gH of extracellular virions is bound to gL and so recognized by gH/gL-specific mAbs; following endocytosis gH/gL epitopes are lost and gH-only epitopes retained; upon fusion , which occurs in late endosomes , gH-only epitopes are also lost . Similarly , the gB of extracellular virions is recognized by mAb BN-1A7 but not mAb MG-1A12; following endocytosis gB gains MG-1A12 recognition; then upon fusion it loses BN-1A7 recognition [45] . Thus virions progress from BN-1A7+MG-1A12−gH/gL+gH-onlylo ( extracellular ) to BN-1A7+MG-1A12+gH/gL−gH-onlyhi ( post-endocytic , but still pre-fusion ) to BN-1A7−MG-1A12+gH/gL−gH-only− ( post-fusion ) . This is illustrated in Fig . 8b: BHK-21 cell-derived virions bound to NMuMG cells at 4°C were not recognized by mAb MG-1A12 ( post-endocytic gB ) and were poorly recognized by mAb MG-9B10 ( gH-only ) ; after endocytosis , virions started to lose recognition by T2C12 and BN-1A7 , strongly gained recognition by MG-1A12 , and showed some increase in recognition by MG-9B10 . The residual BN-1A7 , T2C12 and MG-9B10 staining after 2 h at 37°C was outside LAMP-1+ late endosomes , that is on virions that had not yet reached their site of fusion . RAW-264 cell-derived virions by contrast constitutively showed the post-endocytic forms of gB and gH ( Fig . 8c ) . The mAb used here to detect pre-fusion gB - SC-9E8 - behaves the same as mAb BN-1A7 [45] , and staining for gL followed the same pattern as gH/gL . Thus myeloid-derived virions appeared to penetrate B cells better because they were already primed for membrane fusion . An important in vivo role for myeloid infection in B cell colonization would predict that it occurs early after host entry . We tested this by immunostaining acutely infected noses . At 1 day post-infection F4/80+ macrophages contacted viral eGFP+ epithelial cells ( Fig . 9a ) , and infected macrophages occupied areas of viral lytic gene expression ( Fig . 9b ) . Therefore myeloid infection immediately followed epithelial infection and preceded B cell infection ( Fig . 1 ) . To identify whether the virus reaching B cells had previously replicated in a myeloid cell we infected with the MHV-RG floxed reporter virus lysM-cre mice , which express cre recombinase in macrophages , granulocytes , and some dendritic cells [46] ( Fig . 9c ) . Virus recombination rates varied between mice , but at 3 days post-inoculation approximately 1/3 of the MHV-RG recovered from lysM-cre mouse noses was eGFP+mCherry− , indicating prior replication in a myeloid cell . At days 5 and 8 the mean recombination rate was 1/2 . By contrast almost all the virus recovered from noses of CD19-cre mice , whch express cre recombinase in B cells [47] , remained eGFP−mCherry+ , consistent with B cells not being a primary infection target in the upper respiratory tract , and with myeloid infection preceding that of B cells . B cells are the major site of MuHV-4 lymphoid infection [29] , and virus recovered from the draining lymph nodes ( SCLN ) of CD19-cre mice was 80% eGFP+mCherry− at day 8 and almost 100% eGFP+mCherry− at day 15 . LysM-cre mice again showed individual variation , but on average more than half the SCLN virus was eGFP+mCherry− . At day 15 post-infection we also assayed CD11c-cre mice , which express cre recombinase predominantly in dendritic cells [48] . Virus recovered from their SCLN was >90% eGFP+mCherry− . LysM and CD11c expression show only modest overlap [46] . Thus most if not all the virus reaching lymph nodes and infecting B cells appeared to have passed through at least one myeloid cell .
An enduring puzzle with MuHV-4 and KSHV has been that they persist in B cells in vivo yet infect them poorly in vitro . For MuHV-4 this is despite virions growing to high titers and efficiently infecting mice . We identified a binding block to myeloid infection that was overcome by co-culture , and a post-binding block to B cell infection that was overcome by virus propagation in myeloid cells . Better B cell penetration was associated with virions displaying a post-endocytic form of the gB/gH fusion complex . Consistent with these tropisms , host entry was an epithelial infection; myeloid infection followed; lymphoid infection occurred only later; and most if not all of the virus recovered from lymph nodes showed previous myeloid infection . Virion tropism therefore matched host antigen transport in setting an epithelial to myeloid to lymphoid infection cycle . Virus binding also presented a hurdle to B cell infection . This increased when B cell HS expression increased . However while plasma cell differentiation upregulates B cell HS [49] , MuHV-4 persists in memory rather than plasma B cells; and while interferon upregulates B cell HS [50] it also inhibits MuHV-4 infection . Therefore in vivo B cell binding seems unlikely to depend on HS up-regulation . Binding also increased when virions lacked gp150 , but whether such virions are naturally produced is unclear: in contrast to the Bovine Herpesvirus-4 gp180 [51] , RT-PCR has not demonstrated gp150 truncation by splicing ( unpublished data ) ; nor were myeloid cell-derived virions functionally gp150-deficient . The relative gp150-independence of B cell infection by co-culture suggested that binding occurs instead through myeloid/B cell contact [52] . Myeloid infection in turn probably involves contact with infected epithelial cells . Thus a need for HS binding by cell-free virions applies mainly to host entry [53] , where the target was epithelial . Penetration presented a more severe block to B cell infection . B cell penetration by EBV requires its gp42 [54] . MuHV-4 and KSHV lack obvious gp42 homologs and no MuHV-4 glycoprotein knockout has specifically failed to infect B cells [24] , [40] , [53] , [55]–[57] . Therefore a B cell-specific component to the MuHV-4 fusion complex , while difficult to exclude , seems unlikely . Instead B cell-tropic virions showed antigenic changes in gB and gH . These corresponded to post-endocytic conformation changes that normally precede epithelial membrane fusion [45] . If these conformation changes are a prerequisite for fusion and are not triggered by B cells , it may be crucial that they are triggered by virion exit through an endocytic/exocytic compartment of myeloid cells . Simple co-culture with macrophages increased B cell infection at least 100-fold; specific lymphoid architecture and myeloid/B cell communication pathways doubtless make the process more efficient still in vivo . Most in vivo myeloid populations are heterogeneous , but classically dendritic cells move from peripheral sites to lymph nodes , whereas macrophages are sessile . Thus dendritic cells could play an important role in virus transport . Alternatively , lymphatic virion transport could connect roles for macrophages in the periphery , where F4/80+ cells were closely associated with epithelial infection , and in lymph nodes , where viral DNA is found in CD11b+CD11c− cells [18] . CD11c+ cells are mainly dendritic and lysM+ cells mainly macrophages . Some macrophages express CD11c [58] , [59] and some CD11c+ cells express lysM , but the overlap is not extensive and not even all macrophages express lysM [46] . Thus the high percentages of MHV-RG recombination in both CD11c-cre and lysM-cre mice suggested that MuHV-4 might infect both macrophages and dendritic cells before reaching B cells . The epithelial/myeloid/lymphoid MuHV-4 infection pathway is quite different to the epithelial cell/B cell exchange proposed for EBV [60] . Thus despite rhadinoviruses and lymphocryptoviruses colonizing similar cell populations , they may do so in different ways . Epithelial and fibroblast-derived MuHV-4 are strongly epithelial-tropic - we find little difference between epithelial and fibroblast infections [55] - as are NS0 cell-derived virions ( data not shown ) ; and even RAW-264 cell-derived MuHV-4 infected epithelial cells better than B cells . This consistent epithelial tropism perhaps reflects a predominant need for epithelial entry in the viral lifecycle: new infected B cells can come from lymphoproliferation , but each new epithelial infection likely requires new cell binding and penetration . EBV may have evolved ways to minimise its need for new epithelial infections . However it does not present clinically until infection is well established , so events equivalent in timing to the epithelial and myeloid infections of MuHV-4 are rarely studied . The striking parallels between MuHV-4 host colonization and normal antigen transport would suggest that other lymphotropic viruses follow similar routes .
All animal experiments were approved by the University of Cambridge ethical review board and by the UK Home Office under the 1986 Animal ( Scientific Procedures ) Act as Project Licence 80/2538 . C57BL/6 ( Harlan UK ) , LysM-cre [46] , CD19-cre [47] and CD11c-cre mice [48] were housed in the Cambridge University Department of Pathology animal unit . LysM-cre mice express cre recombinase in monocyte/macrophages and neutrophils in place of myeloid-specific lysozyme , CD19-cre mice express cre in B cells , and CD11c-cre mice express cre predominantly in dendritic cells . Mice were given MuHV-4 intranasally ( i . n . ) either in 30 µl under general anesthesia to infect both the upper and lower respiratory tract , or in 5 µl without anesthesia to infect just the upper respiratory tract . BHK-21 fibroblasts ( American type culture collection CCL-10 ) , CHO-K1 epithelial cells ( CCL-61 ) , the glycosaminoglycan-deficient CHO-745 mutant ( CRL-2242 ) , NMuMG epithelial cells ( CRL-1636 ) , NS0 myeloma cells , A20 B cells ( TIB-208 ) , NIH-3T3 cells ( CRL-1658 ) , NIH-3T3-cre cells [61] 293T cells ( CRL-11268 ) , and RAW-264 monocytes ( TIB-71 ) were cultured in Dulbecco's Modified Eagle's Medium with 2 mM glutamine , 100 U/ml penicillin , 100 mg/ml streptomycin and 10% fetal calf serum . Primary cells were cultured in the same medium supplemented with 50 mM 2-mercaptoethanol . Peritoneal macrophages were harvested 5 days after intraperitoneal injection of Brewer's thioglycollate medium ( Sigma Chemical Co . ) , by injecting and aspirating post-mortem 10 ml of Dulbecco's Modified Eagle's Medium . Cells non-adherent to tissue culture plates ( Nunc ) after 2 h at 37°C were discarded . The remaining cells were routinely >90% CD11bhiCD19− by flow cytometry . Spleens were disrupted into single cell suspensions by homogenization in a Griffiths tube . Debris was removed by filtration ( 200 µm ) , and erythrocytes and dead cells were removed by centrifugation on Ficoll . The cells recovered were typically 60% B cells CD19+ , 25% CD4+ T cells and 15% CD8+ T cells . <1% of the recovered cells were macrophages ( CD11bhi ) or dendritic cells ( CD11c+ ) . Retroviral transduction of A20 cells with an uncleavable form of the syndecan-1 extracellular domain has been described [30] . We used the same approach to over-express syndecan-1 in RAW-264 cells , selecting transduced cells with Zeocin ( Invitrogen ) . As with NIH-3T3-cre cells , we generated NS0-cre cells by transduction with a cre expressing retrovirus and selected transduced cells with G418 . All viruses were generated from a BAC-cloned MuHV-4 genome [62] . Gp150+ and gp150− versions of MuHV-4 with a C-terminal eGFP tag on the abundant virion envelope component gM have been described [38] , as has the generation of MuHV-4 with an intergenic EF1α-eGFP expression cassette [63] . We made a gp150− version of the EF1α-eGFP BAC by RecA-mediated recombination of a genomic clone containing multiple stop codons at genomic co-ordinate 69743 [24] . This terminated the 483 amino acid gp150 coding sequence after 93 amino acids . To make MuHV-4-RG , in which cre recombinase switched reporter gene expression from mCherry to eGFP , we started with a derivative of pEGFP-C2 ( Clontech ) in which the eGFP coding sequence had been replaced by that of mCherry [64] . We excised most of the polylinker by digesting with BamHI + BglII , gel purifying , and religating , then PCR-amplified the mCherry coding sequence plus the downstream SV40 polyadenylation site with primers adding an outer EcoRI restriction site and an altered loxP site ( loxP* ) to each flank . The loxP* spacer region was GGATACTT rather than GCATACAT to make it incompatible with the loxP sites flanking the MuHV-4 BAC cassette . EGFP expression from an intergenic MuHV-4 M3 promoter ( pM3-eGFP-pA ) has been described [65] . We cloned the EcoRI-restricted loxP*-mCherry-pA-loxP* PCR product into the EcoRI site between the M3 promoter and the eGFP start codon of pM3-eGFP-pA in pSP73 . Genomic flanks for recombination were then added by blunt end cloning pM3-loxP*-mCherry-pA-loxP*-eGFP-pA into the MfeI site ( genomic coordinate 77176 ) of a BglII genomic clone ( 75338–78717 ) in pSP73 , then subcloning as a BglII fragment into the BamHI site of the KanR+SacB+orits shuttle vector pST76K-SR . pM3-loxP*-mCherry-pA-loxP*-eGFP-pA was then recombined into the MuHV-4 BAC by transient RecA expression , selection with kanamycin , and counter-selection with sucrose [62] . Recombinant clones were identified and checked for genomic integrity by restriction enzyme mapping . Infectious virus ( mCherry+ from the reporter construct and eGFP+ from the BAC cassette ) was recovered by transfecting BAC DNA into BHK-21 cells . This was then passed once through NIH-3T3-cre cells ( infection at 2 p . f . u . /cell ) . Viruses excising the BAC cassette but retaining the complete loxP*-flanked reporter cassette ( eGFP−mCherry+ ) were selected from the mixed progeny by flow cytometric sorting of infected cells and cloning on BHK-21 cells . Correct insertion of the expression cassette was confirmed by viral DNA sequencing . Viruses were grown in BHK-21 cells by low multiplicity infection ( 0 . 01 p . f . u . /cell ) and culture until >50% of cells showed cytopathic effects - typically 3–5 days . Viruses were grown in RAW-264 cells , which support lytic propagation less well , by infection at 0 . 1 p . f . u . /cell and culture for 2–3 weeks , collecting the medium every 3–4 days and sub-culturing the cells as required . By this time >50% of the cells showed cytopathic effects . Virions were harvested from infected cell supernatants by ultracentrifugation ( 35 , 000× g , 90 min ) . Cell debris was removed by low speed centrifugation ( 500× g , 10 min ) and by filtration ( 0 . 45 µm ) . Expression constructs for the N-terminal 3 short consensus repeats of gp70 and the N-terminal 250 amino acids of gp150 , each fused to human IgG-Fc , have been described [37] . These were transfected into 293T cells using Fugene-6 ( Roche Diagnostic Ltd ) . Recombinant proteins were collected from cell supernatants . The amounts of each Fc fusion were quantitated by immunoblotting for IgG-Fc and normalized on this basis . Tissue blocks containing the whole upper respiratory tract epithelium [64] were homogenized in a pestle and mortar; lungs were homogenized mechanically ( Omni International ) . Virus titers were then determined by plaque assay [24] . BHK-21 cell monolayers were incubated with virus dilutions ( 2 h , 37°C ) , overlaid with 0 . 3% carboxymethylcellulose and 4 days later fixed with 4% formaldehyde and stained with 0 . 1% toluidine blue for plaque counting . Virus titers in spleens and lymph nodes were determined by infectious centre assay of single cell suspensions [24] . Tissue samples from mice infected with MHV-RG were plated at limiting dilution and positive wells scored for red or green fluorescence under UV illumination . Viral genome loads were measured by Q-PCR [64] . MuHV-4 genomic co-ordinates 4166–4252 ( M2 gene ) was amplified ( Rotor Gene 3000 , Corbett Research ) from 50 ng DNA extracted from ex vivo organs ( Promega Corporation ) . The PCR products were quantitated by hybridization with a Taqman probe ( genomic coordinates 4218–4189 ) and converted to genome copies by comparison with a standard curve of cloned plasmid template , amplified in parallel . Cellular DNA was quantitated in the same reaction by amplifying part of the adenosine phosphoribosyl transferase ( APRT ) gene , again with Taqman probe hybridization and known template dilutions amplified in parallel for quantitation . Virus loads were then normalized by the cellular genome copy number of each sample . Green ( eGFP ) and red ( mCherry ) fluorescence were measured directly . For antibody staining , adherent cells were plated overnight onto Petri dishes , then detached without trypsinization . Non-adherent cells were used directly . B cells were identified by staining with a phycoerythrin-conjugated rat mAb to CD19 ( BD Biosciences ) or an Alexafluor633-conjugated pAb to mouse immunoglobulin ( Invitrogen ) . Syndecan-1 and syndecan-4 were detected with phycoerythrin-conjugated mAbs ( BD Biosciences ) ; heparan sulfate was detected with mAbs F58-10E4 and NAH46 ( Seikagaku Corporation ) plus Alexafluor488-conjugated pAb to mouse immunoglobulin ( Invitrogen ) . Analysis was performed on a FACS Calibur and sorting on a FACS Vantage ( BD Biosciences ) . Virions were denatured by heating in Laemmli's buffer ( 50°C , 10 min ) , then proteins resolved by SDS-PAGE and either stained with Coomassie Brilliant Blue followed by destaining in acetic acid/methanol , or transferred to PVDF membranes . The membranes were blocked in 10% non-fat milk then incubated with mAbs specific for the gB N-terminus ( MG-2C10 ) [41] , the C-terminal half of gB ( MG-4D11 ) [43] , gp70 ( 9C7 ) [33] , gN ( 3F7 ) [66] , gp150 ( T1A1 ) [24] and thymidine kinase ( CS-4A5 ) [67] . EGFP was detected with a rabbit pAb ( Abcam ) . Antibody binding was detected with horseradish peroxidase-conjugated rabbit anti-mouse IgG pAb or donkey anti-rabbit IgG pAb ( Dako Corporation ) . Development was with ECL reagents ( APBiotech ) and exposure to X-ray film . To analyse virion antigenicity , NMuMG cells were adhered to glass coverslips . Virions were then added to the cells at 4°C to allow binding but not endocytosis . The cells were then fixed ( 2% formaldehyde , 30 min , 4°C ) and blocked in PBS/5% fetal calf serum/0 . 1% Tween-20 . Pre-endocytic gB was detected with mAbs SC-9E8 or BN-1A7 [45] , post-endocytic gB with mAb MG-1A12 [43] , gH/gL with mAb T2C12 [55] , gL with mAb 47-5G10 [68] and gH-only with mAb MG-9B10 [55] . In some experiments LAMP-1 was detected with mAb 104B ( BD Biosciences ) . After the primary antibody incubations ( 1 h , 23°C ) the cells were washed ×3 in PBS/0 . 1% Tween-20 , incubated with Alexafluor 568-coupled goat anti-mouse-IgG pAb or with Alexafluor 488-coupled goat anti-mouse-IgG pAb + Alexafluor 568-coupled goat anti-rat IgG pAb ( 1 h , 23°C ) ( Invitrogen ) , washed ×3 in PBS/0 . 1% Tween-20 , mounted in Prolong Gold with DAPI ( Invitrogen ) , and imaged with an Olympus microscope plus Hamamatsu digital camera or with a Leica SP2 confocal microscope . The anterior part of the skull containing the olfactory epithelium was removed post-mortem and fixed in 4% formaldehyde–PBS ( 4°C , 24 h ) . Samples were decalcified in 250 mM EDTA ( two weeks , 23°C , changing the buffer every 2–3 days ) , then washed ×2 in PBS and paraffin-embedded . 7 µm sections were cut , de-paraffinised in xylene and rehydrated in ethanol/water . Antigen retrieval was performed by microwaving ( 850W , 5 min ) in 10 mM NaCitrate pH 6/0 . 05% Tween-20 . Endogenous peroxidase activity was quenched in PBS/3% H2O2 for 10 min . The sections were then blocked with 2% rabbit serum and viral antigens detected with a MuHV-4-immune rabbit serum ( 18 h , 23°C ) [27] , biotinylated goat anti-rabbit IgG pAb and Vectastain Elite ABC Peroxidase system with ImmPACT DAB substrate ( Vector Laboratories ) , washing ×3 in PBS between each step . The sections were counterstained with Mayer's Hemalum ( Merck ) and mounted in DPX ( BDH ) . Viral miRNA/tRNAs 1–4 were detected by in situ hybridisation [28] . After de-waxing in xylene and rehydration in ethanol/water , fixed sections were treated with proteinase K ( 100 µg/ml , 10 min , 37°C ) and acetylated with 25% acetic anhydride in 0 . 1 M triethanolamine . They were then hybridized in 50% formamide/10 mM Tris pH 7 . 5 with a digoxigenin-labelled riboprobe , generated by T7 transcription of pEH1 . 4 ( 58°C , 18 h ) . Hybridized probe was detected with alkaline phosphatase-conjugated anti-digoxigenin Fab fragments ( Boehringer Ingelheim ) and BCIP/NBT substrate . For fluorescence imaging , samples were fixed in 1% formaldehyde/10 mM sodium periodate/75 mM L-lysine ( 4°C , 24 h ) , equilibrated in 30% sucrose ( 4°C , 18 h ) , then frozen in OCT and sectioned ( 7 µm ) on a cryostat . Sections were air dried ( 2 h , 23°C ) and blocked with 2% serum/2% BSA/PBS ( 1 h , 23°C ) . We detected macrophages with mAb F4/80 ( Serotec ) plus Alexafluor568-conjugated goat anti-rat IgG pAb ( Invitrogen ) . We detected viral eGFP expression with rabbit anti-eGFP pAb ( Abcam ) plus Alexafluor488-conjugated goat anti-rabbit IgG pAb ( Invitrogen ) . Sections were washed ×3 in PBS after each antibody incubation ( 1 h , 23°C ) , then mounted in Prolong Gold + DAPI ( Invitrogen ) , visualised using a Leica TCS SP2 confocal microscope , and analysed with ImageJ . | Rhadinoviruses cause lymphocytic cancers . Their infection of lymphocytes is therefore an important therapeutic target . How this occurs is unclear . One prevalent hypothesis has been that virions directly infect lymphocytes when they enter new hosts . Here we show that host entry by Murid Herpesvirus-4 , a close relative of the Kaposi's Sarcoma-associated Herpesvirus , is an epithelial rather than a lymphocyte infection: the mucosal lymphoid colonization typical of acute infectious mononucleosis only occurred later . Macrophages were closely associated with the acutely infected epithelium , and most if not all of the virus reaching B cells showed evidence of previous myeloid cell infection . Macrophage-derived virions showed a greatly enhanced capacity for lymphocyte infection that was associated with antigenic changes in the viral fusion proteins . Thus host colonization required epithelial and myeloid infections before there was lymphocyte infection . The implication is that each of these infection events could be independently targeted to limit viral persistence . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"models",
"of",
"infection",
"viral",
"entry",
"host",
"cells",
"viral",
"persistence",
"and",
"latency",
"viral",
"transmission",
"and",
"infection",
"virology",
"biology",
"microbiology",
"host-pathogen",
"interaction"
] | 2012 | Myeloid Infection Links Epithelial and B Cell Tropisms of Murid Herpesvirus-4 |
The dopamine D2 and D3 receptors ( D2R and D3R ) are important targets for antipsychotics and for the treatment of drug abuse . SB269652 , a bitopic ligand that simultaneously binds both the orthosteric binding site ( OBS ) and a secondary binding pocket ( SBP ) in both D2R and D3R , was found to be a negative allosteric modulator . Previous studies identified Glu2 . 65 in the SBP to be a key determinant of both the affinity of SB269652 and the magnitude of its cooperativity with orthosteric ligands , as the E2 . 65A mutation decreased both of these parameters . However , the proposed hydrogen bond ( H-bond ) between Glu2 . 65 and the indole moiety of SB269652 is not a strong interaction , and a structure activity relationship study of SB269652 indicates that this H-bond may not be the only element that determines its allosteric properties . To understand the structural basis of the observed phenotype of E2 . 65A , we carried out molecular dynamics simulations with a cumulative length of ~77 μs of D2R and D3R wild-type and their E2 . 65A mutants bound to SB269652 . In combination with Markov state model analysis and by characterizing the equilibria of ligand binding modes in different conditions , we found that in both D2R and D3R , whereas the tetrahydroisoquinoline moiety of SB269652 is stably bound in the OBS , the indole-2-carboxamide moiety is dynamic and only intermittently forms H-bonds with Glu2 . 65 . Our results also indicate that the E2 . 65A mutation significantly affects the overall shape and size of the SBP , as well as the conformation of the N terminus . Thus , our findings suggest that the key role of Glu2 . 65 in mediating the allosteric properties of SB269652 extends beyond a direct interaction with SB269652 , and provide structural insights for rational design of SB269652 derivatives that may retain its allosteric properties .
G protein-coupled receptors ( GPCRs ) represent one of the largest protein families , and regulate a myriad of physiological processes in response to diverse chemical or environmental stimuli [1] . Among this family , members of the dopamine D2-like receptor subgroup ( consisting of dopamine D2 receptor ( D2R ) , D3R , and D4R ) have been implicated in various physiological functions , including voluntary movement , reward , sleep , learning , and memory [2] . Previous studies have established dopamine D2-like receptors as important therapeutic targets for a variety of neuropsychiatric disorders as well as for the treatment of drug addictions [2 , 3] . Over the last two decades , significant efforts have been made towards understanding the structure-function relationships of these receptors [4–6] . Despite this success , the high sequence identity within the subgroup presents a formidable challenge for selective drug development [7] . In recent years , several bitopic ligands that target both the orthosteric binding site ( OBS ) and a secondary “allosteric” binding site in GPCRs have been developed to achieve subtype specificity , improve binding affinity , and lead to a reduction in the side effects compared to orthosteric ligands [8] . Whereas most bitopic ligands show competitive behavior against other ligands that bind the OBS [8] , SB269652 , a bitopic ligand for D2R and D3R , has been shown to act as an allosteric modulator at both receptors [9–12] . SB269652 is composed of a tetrahydroisoquinoline ( THIQ ) and an indole-2-carboxamide moiety , connected by a cyclohexyl linker in trans orientation . Molecular modeling of SB269652 in D2R showed that the THIQ moiety binds in the OBS and forms an ionic interaction with Asp3 . 32 ( superscripts denote Ballesteros-Weinstein numbering [13] ) , while the indole-2-carboxamide moiety protrudes into a secondary binding pocket ( SBP ) formed by the extracellular portions of transmembrane segments ( TMs ) 2 and 7 . The pose in the SBP establishes a hydrogen bond ( H-bond ) between the N atom of the indole-2-carboxamide and Glu2 . 65 [10] . An N-methyl indole-2-carboxamide derivative of SB269652 that is no longer able to make this interaction displayed competitive behavior [14] , consistent with an alteration in the binding of the ligand in the SBP . Derivatives based on the indole-2-carboxamide moiety , N-isopropyl-1H-indole-2-carboxamide and N-butyl-1H-indole-2-carboxamide , were recently found to display allosteric pharmacology in D2R and D3R , respectively [12 , 15] , which suggest that the SBP near TMs 2 and 7 is indeed an allosteric binding site . In addition , SB269652 was inferred to mediate negative allosteric modulation through a dimer interface of D2R [10] . Mutagenesis experiments implicated Glu2 . 65 near the proposed TM1 dimer interface of D2R [16] as a key determinant for the activity of SB296652 , as replacement of this residue with alanine caused a decrease in both SB269652 affinity and negative cooperativity [10] . Similar disruption by the E2 . 65A mutation of SB269652 binding affinity was also observed at D3R . However , as the proposed H-bond between Glu2 . 65 and the indole moiety of SB269652 is not a strong interaction and the E2 . 65A mutation did not change the pharmacological profile of SB269652 from allosteric to competitive , the H-bond may not be the only element to determine the allosteric properties [10 , 14 , 15] . Indeed , our structure activity relationship ( SAR ) studies also suggested that the size and lipophilicity of the indole-2-carboxamide moiety were also determinants of allosteric pharmacology [14] . Thus , another impact of the E2 . 65A mutation , such as the potentially altered size and shape of the SBP in response to the mutation , may also contribute to the decrease in affinity and negative cooperativity . In the present study , we carried out extensive molecular dynamics ( MD ) simulations to characterize differences in the binding modes of SB269652 in D2R or D3R , and the impact of the E2 . 65A mutation . Our results elucidate important mechanistic details of the role of Glu2 . 65 in the SBP-mediated change in binding affinity and negative cooperativity .
We carried out comparative MD simulations of four conditions: D2R and D3R wild-type ( WT ) and their E2 . 65A mutants bound to SB269652 . The D2R models in complex with SB269652 were derived from our previous study [10] , whereas the starting poses of SB269652 in our D3R models are similar to those in D2R models ( see Methods ) . The first set of simulations was followed by multiple rounds of additional simulations , in which we collected more trajectories for the under-sampled microstates based on the results of the Markov state model ( MSM ) analysis [17 , 18] ( see Methods ) . In total , we collected 145 MD trajectories with a cumulative length of 76 . 5 μs ( Table 1 ) . Similar to our previous study [10] , in the resulting conformations from our extensive MD simulations , the primary pharmacophore ( PP ) of SB269652 , the THIQ moiety , forms a salt bridge with the carboxyl group of Asp3 . 32 in both D2R and D3R , a key component of ligand binding to aminergic receptors [7] . The secondary pharmacophore ( SP ) , which consists of an indole-2-carboxamide moiety , attached to the PP through a trans-cyclohexylene linker , shows significant dynamics in all our simulated conditions ( Fig 1 ) . To characterize the dynamics of the SP poses , we performed MSM analysis to identify the thermodynamic populations of the SB269652 binding poses and to calculate the kinetics of transitions between these populations . Specifically , we used 12 distances between the nitrogen atoms of SB269652 and the Cβ atoms of selected binding-site residues and 4 intra-ligand measures as the input features ( see Methods and S1 Fig ) . The analysis identified two metastable states ( MSs ) with similar equilibrium probabilities of 52% and 48% in D2R/WT ( shown as orange and green spheres in Fig 1C ) . In the green MS , SB269652 forms two H-bonds to Glu2 . 65 with both the indole N4 and amide N3 atoms as we described previously [10] ( Fig 2A and 2C ) . However , in the orange MS , the H-bond between the N4 atom and Glu2 . 65 is lost as N4 reorients toward the extracellular side ( Fig 1C ) . In addition , whereas N3 is in a similar orientation as in the green MS , it has significantly reduced propensity to form a H-bond with Glu2 . 65 ( Fig 2A and 2C ) . In contrast to the observed dynamics of the SP among different MSs , in both green and orange MSs , the PP is stable and the salt-bridge interaction between the charged N1 nitrogen in the PP and the key binding-site residue Asp3 . 32 remained intact ( S2A Fig ) , suggesting that the strong salt-bridge interaction deters the dynamics of the SP from propagating to the PP , although we have found that the poses of the PP and SP of bivalent ligands can affect each other [10 , 20 , 21] . For D3R/WT , we found that the two states identified by the MSM analysis are similar to those in D2R/WT , in terms of the orientations of the indole-2-carboxamide moiety of SB269652 , relative to Glu2 . 65 . Interestingly , in D3R/WT the orange MS in which the N4 atom of SB269652 faces toward the extracellular side also has a slightly higher equilibrium probability ( 53% ) than the green MS ( 47% ) with the N4 atom interacting with Glu2 . 65 ( Fig 1D ) . Similar to D2R/WT , the PP is stable in D3R/WT as well , with an intact interaction between the N1 nitrogen and Asp3 . 32 ( S2B Fig ) . Although the PP is stable in both D2R and D3R , we observed subtly different poses in the OBS of these two receptors . Comparing the representative poses of SB269652 at D2R and D3R , we noticed different interactions between the THIQ moiety and residues from extracellular loop 2 ( EL2 ) . Specifically , the subtle divergence of these two receptors at the interface between EL2 and EL1-TM2 accommodates the cyclohexyl linker of SB269652 slightly differently , and this divergence appears to correlate with drastically different orientations of the conserved Ile at the EL2 . 52 position ( second residue after the conserved disulfide-bonded Cys in EL2 ) : while Ile183EL2 . 52 in D3R forms a favored hydrophobic interaction with the THIQ moiety in the OBS , Ile184EL2 . 52 in D2R points upwards and is not in contact with SB269652 ( Fig 3 and S1 Table ) . Such a difference is consistent with the results of our per-residue decompositions of the MM/GBSA binding energy calculations of the representative D2/WT and D3/WT frames , in which IleEL2 . 52 contributed favorably to binding of SB269652 at D3R but not at D2R . In addition , we found that SB269652 interacts with Ser1935 . 42 , Ser1945 . 43 , and Ser1975 . 46 in D2R , while it only interacts with Ser1925 . 42 in D3R . This is likely due to the divergence in both EL2 and TM5 between D2R and D3R –in addition to the divergent EL2 . 51 and EL2 . 53 positions in EL2 , TM5 is divergent at position 5 . 52 ( Ile2035 . 52 in D2R and Gly2025 . 52 in D3R ) near the proline5 . 50-induced kink ( Fig 3 ) . Previously , it was found that SB269652 had more than 10-fold higher binding affinity at D3R than at D2R , and a chimera mutagenesis study that swapped the D2R and D3R segments identified EL2 and TM5 to be important for the different binding affinities [9] . Thus , our findings of the divergent poses of SB269652 in the OBS of D2R and D3R are highly consistent with these results . In comparison to D2R/WT , our MSM analysis identified 3 MSs for the D2R/E2 . 65A condition . One MS of D2R/E2 . 65A is similar to the green MS of D2R/WT; however , given the absence of the H-bond between the indole-2-carboxamide moiety and Ala2 . 65 , the indole ring of SB269652 in the green MS of D2R/E2 . 65A tends to be more parallel to the membrane compared to in D2R/WT ( Fig 1E ) . In the dominant new pose of the SP in the D2R/E2 . 65A condition ( magenta MS in Fig 1E , which has an equilibrium probability of 60% ) , both the amide N3 and indole N4 atoms face toward the extracellular side , but the amide O atom faces the intracellular side , which is rarely observed in D2R/WT ( Fig 1C and 1E ) . Interestingly in D3R/E2 . 65A , the three MSs we identified ( Fig 1F ) have significant similarity to those three in the D2R/E2 . 65A , in terms of the distances of N3 and N4 to Ala2 . 65 ( S3 Fig ) . Even though the orange MS is the most dominant MS ( 69% ) in D3R/E2 . 65A instead of the magenta MS in D2R/E2 . 65A ( Fig 1F ) , in both mutant receptors , N4 of SB269652 has a similar tendency to face away from Ala2 . 65 . We hypothesized that in addition to the H-bonds between the indole-2-carboxamide moiety of SB269652 and Glu2 . 65 , another key to understanding the significance of the E2 . 65A mutation on the allosteric action of SB269652 lies in conformational changes resulting from this mutation . Our structural analysis identified marked conformational differences between the D2R/WT and D2R/E2 . 65A conditions bound with SB269652 , in the SBP consisting of TM1e , TM2e , TM3e , and TM7e subsegments ( see S2 Table for the division of subsegments [20 , 22] ) . These differences were characterized by a significantly larger TM2e-TM7e distance and a shorter TM1e-TM3e distance in D2R/E2 . 65A as compared to the D2R/WT condition , demonstrating the altered size and shape of the SBP in the mutant construct ( Fig 4 ) . Interestingly , the occupation of the SBP by the SP of SB269652 in D2R/WT increased both TM2e-TM7e and TM1e-TM3e distances ( Fig 4C ) compared to the D2R/WT condition equilibrated with eticlopride ( Fig 4B and 4D ) , a ligand that predominantly occupies the OBS and does not protrude into the interface between TMs 2 and 7 . Thus it appears that the SBP is dynamically formed to accommodate the SP of SB269652 , and that the E2 . 65A mutation ablates the ability of SB269652 to increase the distance of TM1e-TM3e through the interaction of its SP with the SBP . A similar enlargement of the SBP by SB2696952 was observed for D3R/WT as well ( Fig 4C and 4D ) . Comparing the two D3R conditions bound with SB269652 , the E2 . 65A mutation results in larger TM2e-TM7e and smaller TM1e-TM3e distances , similar to the observations for the D2R ( Fig 4C ) . In both D2R and D3R , Glu2 . 65 of TM2e faces Ser7 . 36 of TM7e , and we found that the disruption of this polar interface by the E2 . 65A mutation contributes to the larger TM2e-TM7e distances . However , Glu2 . 65 and Ser7 . 36 have a significant probability to form a H-bond in D3R/WT ( green MS , 54 . 7±0 . 6%; orange MS , 51 . 7±1 . 8% , for the dataset used in Fig 2 ) but not in D2R/WT ( green MS , 14 . 3±0 . 5%; orange MS , 9 . 0±1 . 3% ) . Thus , the TM2e-TM7e distance appears to be larger in both the D2R/WT and D2R/E2 . 65A conditions than in D3R/WT and D3R/E2 . 65A ( Fig 4C ) , likely due to the shorter EL1 in D2R , consistent with our previous observations and with differences in the tendencies of this interface to accommodate the SP of the bitopic ligands [23] . The impact of E2 . 65A on the SBP is associated with altered conformations of N terminus ( NT ) as well . While the NT always adopts flexible loop conformations in our simulations , our loop clustering analysis ( see Methods ) indicates clearly distinct equilibria and preferences of the loop conformations in different conditions . For the combined analysis of D2R/WT and D2R/E2 . 65A , we found that the mutation significantly shifts equilibrium of the NT conformation towards one of the two most populated clusters shown in WT , and has ~70% occupancy for the dominant magenta MS of D2R/E2 . 65A ( Fig 5 and S3 Table ) . Thus , the NT appears to be more dynamic in D2R/WT and adopts multiple conformations , whereas the E2 . 65A mutation reduces such dynamics . Similarly , we found the most populated cluster in D3R/E2 . 65A has a significant higher population and is significantly different from that of D3R/WT ( Fig 5 , S3 Table ) . Interestingly , it appears that residues 9–13 in D3R have a significant tendency to form a helical conformation , whereas in D2R residues 20 and 21 are dominantly in a bend conformation ( S4 Fig ) . In all conditions , the NT bends down and forms a lid over the extracellular vestibule bringing some of the residues in direct contact with the SP of the ligand ( S1 Table ) . Taken together , our data show that the E2 . 65A mutation alters the shape and size of the SBP in both D2R and D3R , which in turn affects the NT conformation .
In recent years , there have been significant advances in the development of allosteric modulators for GPCRs that have high selectivity and novel modes of action . These modulators may lead to therapeutic agents that have fewer side effects [24–26] . One such an example is SB269652 , which acts as a negative allosteric modulator in both D2R and D3R . Whereas our previous studies identified Glu2 . 65 as a critical residue for allosteric modulation of SB269652 [10] , our follow-up SAR study suggests that other elements are also involved in determining the allosteric properties of SB269652 [14] . By carrying out comparative MD simulations in combination with MSM analysis of D2R and D3R WT and their E2 . 65A mutants , we sought to comprehensively characterize the binding poses and dynamics of SB269652 and the impact of E2 . 65A mutation on the size and shape of the SBP . The results of our MD simulations and MSM analysis revealed that in both D2R/WT and D3R/WT , SB269652 has significant probabilities of not forming the H-bonds with Glu2 . 65 , and its SP is in dynamic equilibria between two poses , although they essentially occupy the same space in the SBP in each receptor ( Figs 1 and 2 ) . These results suggest that the direct H-bond interactions between the ligand SP and Glu2 . 65 are not the only factor that governs the allosteric property of SB269652 in either D2R or D3R . Indeed , the mutation E952 . 65A did not cause a switch from allosteric to competitive pharmacology , but rather caused a decrease in the affinity and negative cooperativity of SB269652 [10] . The dynamic equilibria of the binding poses for the allosteric moiety of SB269652 is likely a common feature shared by the binding of other bitopic ligands in allosteric pockets–in many cases , such pockets of GPCRs are located at peripheral regions , more exposed to the water milieu , and may have more dynamic and flexible properties than the OBS . While unlikely to be revealed by crystallography , such dynamic features can be readily identified and characterized by extensive MD simulations in combination with MSM analysis . Our results also indicate that the E2 . 65A mutation significantly alters the dynamic equilibria of the SP of SB269652 in the SBP and results in new poses of the ligand . The new poses ( magenta MSs ) , although not observed in WT for both receptors , are similar to the orange MSs of WT that have the N4 atom facing away from Glu2 . 65 ( Figs 1C and 2A ) . From the perspective of receptor conformation , we found the substitution of the charged and larger Glu2 . 65 residue to a smaller Ala residue significantly affects the packing in the SBP , leading to larger TM2e-TM7e and smaller TM1e-TM3e distances in both receptors ( Fig 4 ) . Thus , we propose that the combined impact from both the removal of the H-bonds and the altered SBP is responsible for the decrease in affinity and cooperativity of SB269652 observed in the E2 . 65A mutants . Given the significant role of the size and shape of the SBP in mediating the allosteric properties of SB269652 , we can envision that some SB269652 derivatives may have allosteric properties even without the capability of forming H-bonds with Glu2 . 65 , as long as they can induce the necessary conformational changes of the SBP . Such conformational changes may be impaired by the E2 . 65A mutation irrespective of whether a ligand has the capacity to H-bond with Glu2 . 65 . Of note , our SAR studies reveal that an N-methyl indole-2-carboxamide derivative of SB269652 ( MIPS1500 ) displayed apparently competitive behavior at D2R/WT , but acted as a negative allosteric modulator of dopamine at D2R/E952 . 65A [10] . While such observations may reflect the inability of this ligand to form a H-bond with Glu2 . 65 as we originally proposed , the addition of a methyl group also adds bulk to the SP . This may change its orientation within the SBP . Thus , by changing the configuration of the SBP , the E952 . 65A mutation may change the orientation of the N-methyl indole-2-carboxamide moiety of MIPS1500 within the SBP causing the ‘gain’ of allosteric pharmacology . Indeed , the size of the SP has also been shown to be an important determinant of the allosteric pharmacology of SB269652 , as derivatives in which the indole moiety was replaced by a pyrrole or proline moiety display apparently competitive pharmacology at the D2R [14] . Such derivatives retain the ability to form an interaction with Glu2 . 65 but lack the lipophilicity and bulk of the indole moiety . Interestingly , our recent mutagenesis studies reveal that , in a similar manner to the N-methyl indole moiety , the pyrrole derivative displays allosteric pharmacology at D2R/E952 . 65A ( Draper Joyce et al . , manuscript in preparation ) . Such observations are consistent with the conformation of the SBP , and the influence of the SP on this conformation , being central to the allosteric pharmacology of SB269652 . Our results also show that the altered size and shape of the SBP in E2 . 65A mutants could bias the NT towards specific conformations , and in the case of D3R , a distinct one from the most populated conformation in the WT . In all conditions , the NT forms a lid over the extracellular vestibule and is in direct interaction with SB269652 , suggesting a previously unappreciated role of the NT in ligand binding at D2R and D3R . Indeed , the functional roles of the NT have been documented recently in a few closely related homologs , including the α1D-adrenergic [27] , 5-HT2B [28] and μ-opioid receptors [29 , 30] . By systematically examining all the available high-resolution crystal structures of class-A GPCRs bound to small compounds , we found 9 structures of 6 receptors showing direct interactions ( within 5 Å of the heavy atoms ) between NT residues and the small-molecule ligands that at least partially occupy the OBS . Interestingly , many of these small-molecule ligands protrude into the interface between TMs 2 and 7 ( S5 Fig ) . Taken together , our findings highlight the key role of the size and shape SBP , which is determined by Glu2 . 65 , in mediating the allosteric properties of SB269652 , and provide structural insights for the rational design of SB269652 derivatives that may retain these allosteric properties .
The binding mode of SB269652 at D2R was investigated based on our previous study [10] . Briefly , to acquire a reference binding mode of the PP ( tetrahydroisoquinoline ( THIQ ) ) of SB269652 in the high-resolution crystal structure of D3R ( PDB code 3PBL [31] ) , THIQ in the protonated form was first docked into the D3R structure with the induced-fit docking ( IFD ) protocol [32] implemented in Schrödinger suite ( release 2016–1 , Schrödinger , LLC: New York , NY ) . The lowest MM/GBSA energy pose from the largest binding mode cluster was selected as a reference pose for the PP of SB269652 at D3R . Assuming that binding modes of THIQ in the near-identical OBSs of D3R and D2R should be similar , we docked the THIQ into the D2R model [20 , 21 , 23 , 31 , 33] and selected a pose that is closest to the THIQ reference pose in the D3R structure . The full-length SB269652 was then docked into the D2R and D3R models by restraining the PP core [21] to the respective THIQ reference poses ( with RMSD tolerance for the heavy-atom restraints of < 2 . 0 Å ) . To investigate the effect of the E2 . 65A mutation , Glu2 . 65 was mutated to Ala in representative frames from equilibrated WT trajectories , and the charge of the system was neutralized by removing a Na+ ion from the water milieu . MD simulations of the receptor–ligand complexes were performed in the explicit water and 1-palmitoyl-2-oleoylphosphatidylcholine ( POPC ) lipid bilayer environment using Desmond MD System ( version 4 . 5; D . E . Shaw Research , New York , NY ) with the CHARMM36 force field [34–37] and TIP3P water model . The ligand parameters were obtained through the GAAMP server [38] , with the initial force field based on CGenFF assigned by ParamChem [39] . The system charges were neutralized , and 150 mM NaCl was added . The average size of the simulation systems was ~110000 atoms . The protein-membrane relaxation was carried out with a protocol modified from that developed by Schrödinger , LLC . Briefly , the initial energy minimization was followed by equilibration with restraints on all protein and ligand heavy atoms in the beginning for 1 ns , then with restraints only on the protein backbone and ligand heavy atoms for 6 ns . For both the equilibrations and the following unrestrained production runs , we used Langevin constant pressure and temperature dynamical system [40] to maintain the pressure at 1 atm and the temperature at 310K , on an anisotropic flexible periodic cell with a constant-ratio constraint applied on the lipid bilayer in the X-Y plane . For each condition , we collected several rounds of multiple trajectories following the procedure described below . The MSM analysis was performed using the PyEMMA program ( version 2 . 3 . 2 ) [41] . For the input featurizer , we chose the features based on the following considerations to describe the interactions and orientations of SB269652 within the receptor binding sites . The polar and charged interactions between the ligand and protein contribute significantly to ligand binding , while these interactions can be more conveniently defined and characterized by simple geometric measures , compared to hydrophobic and aromatic interactions . For SB269652 , the nitrogen atoms are distributed in both the THIQ and indole-2-caboxyamide moieties , so that the dynamics of the entire ligand can be properly characterized using distances between these atoms and protein residues . Therefore , we identified protein residues with their Cβ atoms within 7 . 0 Å of any of the four nitrogen atoms of the SB269652 , and used these Cβ-N distances as input features . To better account for the orientation of the indole-2-caboxyamide moiety of SB269652 , two additional intramolecular distances from the N4 atom of indole ring were included in the features—one to the N3 nitrogen and one to the oxygen of the amide bond . Further , we calculated the vectors from the centers of mass of 5- or 6-member rings of the SP to N4 , and the projections of these vectors on the axis perpendicular to membrane were included to identify the orientation of the indole ring relative to the plane of the membrane . In total , 16 input features were used ( S1 Fig ) . The slow linear subspace of the input coordinates was estimated by the time-lagged independent component analysis ( TICA ) [42 , 43] on the combined data set of D2R/WT , D2R/E2 . 65A , D3R/WT , and D3R/E2 . 65A conditions , and a dimension reduction was achieved by projecting on the 4 slowest TICA components ( which represent 61 . 7% of cumulative kinetic variance ) . k-means clustering was then employed to discretize the simulated subspace and 100 microstates ( MIs ) were obtained . For a range of numbers of MIs ( 50 , 75 , 100 , 150 , 200 , 300 ) , we estimated an MSM for each situation and concluded that the 100-MI MSM performs best based on two analysis . We first calculated scores in terms of variational principle [44 , 45] , using cross-validation [46] as previously described [47] . This analysis showed the variational scores were comparable for small numbers of MI and decreased when the number of MI is larger than 150 . In addition , the 100-MI MSM showed better convergence of the implied time scales ( ITS ) in terms of the lag times ( S6–S8 Figs ) . The discretized combined data set was then divided into individual simulated conditions to estimate Bayesian MSMs [48] . The Bayesian sampling was used to compute statistical uncertainties of 500 transition matrix samples at each lag time . Convergence of ITS for all MSMs was achieved at 96 ns lag time ( S7 Fig ) , which was used to estimate Bayesian MSM for each condition . The PCCA++ method [49] implemented in PyEMMA was then used to stitch the MIs into metastable states ( MSs ) . In the resulting MSMs , 2 MSs were assigned to D2R/WT and D3R/WT conditions each , whereas 3 MSs were assigned to D2R/E2 . 65A and D3R/E2 . 65A conditions each . The identity of common MSs between different conditions was determined based on the number of shared MIs between them . Further structural and kinetic analysis was performed using frames from those MIs that have > 70% probability belonging to their respective MS . In Fig 1 , the transition rate between two MSs is the inverse mean first passage time , which is the expected hitting time of one MS when starting from the other MS . The π value denotes the equilibrium probability of a given MS , which is the probability to be in the MS that remains unchanged in the Markov model as time progresses . The transition rate and equilibrium probability were computed as described previously [50] . The validity of the MSMs was assessed using Chapman-Kolmogorov tests ( S9 and S10 Figs ) which showed that the MSMs estimated at 96 ns were consistent with the simulation data within the 95% confidence interval computed by 500 bootstrapped samples of trajectories . Generally , the Chapman–Kolmogorov test checks if the MSM models estimated at lag time τ can be used to make predictions for the data at longer times kτ within statistical error , i . e . , if Eq ( 1 ) can be satisfied: P ( kτ ) =Pk ( τ ) ( 1 ) where P ( τ ) is the transition matrix estimated from the data at lag time τ ( the Markov model ) , and P ( kτ ) is the transition matrix estimated from the same data at longer lag times kτ . In practice , we use P ( kτ ) and Pk ( τ ) , respectively , to propagate probability starting from one of the metastable states , and measure how much probability ends up in each metastable state [47] . To adequately and efficiently i ) explore the conformational space and ii ) sample the transitions between MSs , we developed an iterative MD sampling protocol to guide the simulations to i ) the less-than-well-sampled regions and ii ) the saddle points on the energy surface that are likely in between MSs . Thus , by taking advantage of the MSM analysis after each round of simulations , we correspondingly identified both i ) the single-frame microstates ( MIs ) , and ii ) the under-sampled MIs ( < 10 frames ) that are in between MSs ( i . e . , the MIs having similar equilibrium probabilities to be in two or more MSs ) , as the starting points for the next round of simulations . For the selected MIs with more than one frames , we select the representative frames from the more advanced stages of MD simulations . The procedure to identify the MIs satisfying the criteria and to select the frame has been automated with an in-house python script . For the MSM-guided simulations , we collected 300 ns for each trajectory . The simulations were considered to have reached convergence until the biggest change in the equilibrium probabilities for the updated MSMs of each condition was < 5% after including data from the new round of the simulations . For the identifications of ligand contact residues shown in Fig 2 , the results are based on 500 Bayesian Markov model samples 3 frames each from those MIs having > 70% probability of belonging to each MS . For each of the MS we identified residues within 5 . 0 Å of ligand ( heavy atom-heavy atom distances ) in D2R/WT and D3R/WT conditions , and the means and standard deviations of three sets of such samplings are shown in S1 Table . For the sub-segment distance calculations shown in Fig 4 , the TMs in both D2R and D3R were divided into subsegments ( extracellular , middle , and intracellular ) as described in [20] ( see S2 Table ) , the results are based on 3000 Bayesian Markov model samples for each condition with the number of samples for each MS proportional to their equilibrium probability , from those MIs having > 70% probability of belonging to each MS . We performed the clustering analysis of the conformations of the N terminus ( NT ) using the same dataset extracted for Fig 2 ( see above ) for each MS in each condition , and combined data sets for one receptor together . The clustering is based on pairwise RMSD of selected NT residues by iteratively excluding the residues with high ( > 5 . 0 Å ) root mean squared fluctuation ( RMSF ) . The final clustering results are based on residues 6–20 and 22–30 for D2R and 2–25 for D3R to perform superimpositions and RMSD calculations . The computed RMSD matrix was then subjected to hierarchical clustering using Ward algorithm implemented in SciPy . The number of clusters for each receptor was determined so that the intra-cluster mean pairwise RMSD for each cluster is within 5 . 0 Å , unless the given cluster has less than 5% of the total frames . The population of each cluster was re-weighted by equilibrium probabilities of the MSs that their members belong to . The means and standard deviations for the three largest clusters are shown in S3 Table . | G protein-coupled receptors ( GPCRs ) are targets of more than 25% of prescription drugs on the market . Due to their critical roles in human physiology , competitive modulation of these receptors has been found to be associated with many undesired side effects . Allosteric modulation holds the promise of retaining normal receptor function and improving selectivity . However , the underlying molecular mechanisms of the allosteric modulation of GPCRs have remained largely uncharted . The dopamine D2-like receptors have been implicated in voluntary movement , reward , sleep , learning , and memory . Based on previous experimental findings , we computationally characterized the binding of a negative allosteric modulator of dopamine D2 and D3 receptors , and revealed the dynamic binding mode of this modulator in a secondary binding pocket ( SBP ) of the receptors . Our results highlight the key role of a Glu in mediating the allosteric properties of the modulator by shaping the dynamically formed SBP , and shed light on rational design and optimization of allosteric modulators of GPCRs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"chemical",
"characterization",
"crystal",
"structure",
"chemical",
"compounds",
"markov",
"models",
"molecular",
"dynamics",
"condensed",
"matter",
"physics",
"neuroscience",
"organic",
"compounds",
"hormones",
"... | 2018 | The E2.65A mutation disrupts dynamic binding poses of SB269652 at the dopamine D2 and D3 receptors |
Many individuals tested for inherited cancer susceptibility at the BRCA1 gene locus are discovered to have variants of unknown clinical significance ( UCVs ) . Most UCVs cause a single amino acid residue ( missense ) change in the BRCA1 protein . They can be biochemically assayed , but such evaluations are time-consuming and labor-intensive . Computational methods that classify and suggest explanations for UCV impact on protein function can complement functional tests . Here we describe a supervised learning approach to classification of BRCA1 UCVs . Using a novel combination of 16 predictive features , the algorithms were applied to retrospectively classify the impact of 36 BRCA1 C-terminal ( BRCT ) domain UCVs biochemically assayed to measure transactivation function and to blindly classify 54 documented UCVs . Majority vote of three supervised learning algorithms is in agreement with the assay for more than 94% of the UCVs . Two UCVs found deleterious by both the assay and the classifiers reveal a previously uncharacterized putative binding site . Clinicians may soon be able to use computational classifiers such as those described here to better inform patients . These classifiers can be adapted to other cancer susceptibility genes and systematically applied to prioritize the growing number of potential causative loci and variants found by large-scale disease association studies .
The BRCA1 gene encodes a large multifunction protein involved in cell-cycle and centrosome control , transcriptional regulation , and in the DNA damage response [1–3] . Inherited mutations in this gene have been associated with an increased lifetime risk of breast and ovarian cancer ( 6–8 times that of the general population ) [4] . There are several thousand known deleterious BRCA1 mutations that result in frameshifts and/or premature stop codons , producing a truncated protein product [5] . In contrast , the functional impact of most missense variants that result in a single amino acid residue change in BRCA1 protein is not known . The Breast Cancer Information Core database ( http://research . nhgri . nih . gov/bic/ ) , a central repository of BRCA1 and BRCA2 mutations identified in genetic tests , currently contains 487 unique missense BRCA1 variants ( April 2006 ) , of which only 17 have sufficient genetic/epidemiological evidence to be classified as deleterious ( Clinically Important ) and 33 as neutral or of little clinical importance ( Not Clinically Important ) . As genetic testing for inherited disease predispositions becomes more commonplace , predicting the clinical significance of missense variants and other UCVs will be increasingly important for risk assessment . Because most UCVs in BRCA1 and BRCA2 occur at very low population frequencies ( <0 . 0001 ) [6] , direct epidemiological measures , such as familial cosegregation with disease , are often not sufficiently powerful to identify the variants associated with cancer predisposition . A promising approach is to supplement epidemiological and clinical analysis of UCVs with indirect approaches such as biochemical studies of protein function and bioinformatics analysis [6–8] . In the future , physicians and genetic counselors may be able to rely on all these sources of information about UCVs when counseling their patients . Previous bioinformatics analysis of BRCA1 UCVs has depended primarily on measures of evolutionary conservation in multiple sequence alignments of human BRCA1 and related proteins from other organisms [9–11] . Two groups have attempted to include information about BRCA1 protein structure . Williams et al . predicted the impact of 25 missense variants in BRCA1′s C-terminal BRCT domains by considering both conservation and location of variant amino acid residues in an X-ray crystal structure [12] . Variants were predicted deleterious if their properties were similar to properties of biochemically characterized deleterious variants in Escherichia coli Lac Repressor and bacteriophage T4 lysozyme . Mirkovic et al . developed a set of hierarchical rules ( Rule-based decision tree ) based on the conservation , variant structural location , and amino acid residue physiochemical properties of 30 deleterious and seven neutral biochemically characterized BRCA1 missense variants [7] . We have developed a novel combination of 16 predictive features that describe conservation , impact of mutation on protein structure , and amino acid residue properties , and used them as input to computational supervised learning algorithms . These algorithms are trained to learn a generic classification of amino acid residue substitutions and positional contexts . The training set is composed of 618 missense variants in the transcription factor TP53 biochemically characterized as functional or nonfunctional in a transactivation assay [13] . TP53 is a tumor suppressor gene that is inactivated in the majority of human cancers . Our validation set is composed of 36 missense variants in BRCA1′s BRCT domains that were biochemically characterized with a transactivation assay [14] . These 36 variants were selected because they occur in individuals from families with breast or ovarian cancer in which no other deleterious mutation in BRCA1 or BRCA2 was found and were functionally tested under the same protocols and conditions , yielding standardized measurements of each variant's transactivation activity with respect to wild-type . We use the validation set to assess the supervised learners and compare them with algorithms based on evolutionarily allowed amino acid residues or empirically derived rules . The algorithms with greatest correlation between assay and computational predictions are the supervised learners Naïve Bayes [15] , Support Vector Machine [16] , and Random Forest [17] . Given a protein X-ray crystal structure , the supervised learning approach can quickly and accurately predict the outcome of our BRCA1 transactivation assay with greater than 94% accuracy on tested missense variants in the BRCT domains . We have applied the best performing supervised learners to blind prediction of the functional impact of 54 UCVs found in BIC and occurring in the BRCA1 BRCT domains . For each of these UCVs , we produce a consensus prediction and , where possible , a molecular explanation for the impact of the variant . Next , we describe the protocol used to train and validate the supervised learning algorithms , the selection of 16 features used to represent each missense variant to the algorithms , implementation details of each algorithm , and performance assessment criteria ( Methods ) . We then show how a combination of sequence- and structure-based features in a supervised learning setting obviates some of the problems with evolutionary analysis and empirically derived rules , providing specific examples of the strengths and weaknesses of each approach ( Results , Discussion ) . We show that two of the variants found to be deleterious by both the assay and the classifiers may be at a previously uncharacterized protein binding site and that electrostatic changes at the site may weaken the interactions of BRCA1 and protein partners that are important for its functions ( Discussion ) . Finally , we discuss the generalizability of our methods to other cancer susceptibility genes and to large-scale disease association studies ( Discussion ) .
We trained four supervised learning algorithms to discriminate between a set of 398 deleterious/nonfunctional and 220 neutral/functional TP53 missense variants , biochemically characterized in a transactivation assay [13] . The variants were downloaded from the IARC TP53 website ( http://www-p53 . iarc . fr ) . We only used variants capable or incapable of activating transcription for all eight of the TP53 promoters tested in the transactivation assay and located in the core DNA binding domain of TP53 . The 36 BRCA1 BRCT missense variants described in our companion paper [14] were used as an independent validation set for the supervised learners . These variants were also classified by sequence-analysis methods based on evolutionarily allowed amino acid residues: Align Grantham Variation Grantham Deviation ( Align-GVGD ) [18] , Sorting Intolerant from Tolerant ( SIFT ) [19] , Ancestral Sequence [9 , 11] , and empirically derived rules encoded in a decision tree ( Rule-based decision tree ) [7] . Each method was evaluated by its agreement with the BRCA1 transactivation assay on the validation set , according to accuracy ( fraction of all variants correctly classified ) , sensitivity or true positive rate ( fraction of all nonfunctional variants correctly classified ) , specificity or true negative rate ( fraction of all functional variants correctly classified ) , Matthews correlation coefficient [20] , and coverage ( fraction of variants for which a prediction was made ) ( Table 1 ) . Matthews correlation coefficient is defined as and ranges from −1 . 0 ( worst ) to 1 . 0 ( best ) . A coefficient of 0 is equivalent to a random prediction , and less than 0 indicates a worse than random prediction . TP is the number of correctly classified nonfunctional variants , TN the number of correctly classified functional variants , FP the number of incorrectly classified nonfunctional variants , and FN the number of incorrectly classified functional variants . For the Naïve Bayes , Support Vector Machine , Random Forest , Decision Tree , Align-GVGD , and SIFT classifiers , we computed a receiver operating characteristic ( ROC ) curve that quantifies the tradeoff between coverage of detected nonfunctional variants ( true positive rate ) and misclassified functional variants ( false positive rate = 1 − specificity ) . ROC analysis was not possible for the Rule-based decision tree and Ancestral Sequence algorithms , which predict the class of a missense variant but do not provide an associated score . The supervised learning algorithms ( Naïve Bayes , Support Vector Machine , Random Forest , Decision Tree ) were trained by associating each amino acid residue substitution in the TP53 training set with 16 carefully selected predictive features ( Table 2 ) . A vector of features for a single substitution is denoted as . The features describe properties of variant and wild-type residues: local structural environment; physiochemical attributes; and evolutionary conservation . To compute the features , we used DSSP ( a program that calculates a variety of geometrical properties for each amino acid residue in a protein structure ) [21] , MODELLER for comparative protein structure modeling [22] , SAM-T2K for protein sequence alignments and hidden Markov models [23] , and in-house PERL code . We began with a core set of 13 features selected by a correlation analysis between features and classes ( functional or nonfunctional ) of the TP53 variants , as described previously [24] ( Table S3 ) . An additional 18 candidate features were evaluated by adding them to the core set and doing 10-fold cross-validation tests of Support Vector Machine performance . Three features were found to improve performance and were added to the optimal feature set; the others were rejected ( Tables S4 and S5 ) . Next we evaluated Support Vector Machine performance with each of the best 16 features held out . In each case , the 10-fold cross-validation test yielded decreased performance . We used X-ray crystal structures from the Protein Data Bank [25] for the BRCA1 BRCT domains ( 1t29 chain A in complex with BACH1 peptide ) [26] and the DNA binding domain of TP53 ( 1kzy ) [27] . We performed in silico mutations on the structures with the MUTATE_MODEL routine of MODELLER ( available as a Python script at http://salilab . org/modeller/wiki/Mutate_model ) . MUTATE_MODEL substitutes the wild-type amino acid residue at a position of interest with a variant amino acid residue , and optimizes the coordinates of the variant's backbone and sidechain atoms with an initial conjugate gradient minimization , molecular dynamics optimization with simulated annealing , and a final conjugate gradient minimization ( E . Feyfant , 2004 , private communication ) . The amino acid residue sequences of human TP53 ( P04637 ) and BRCA1 ( P38398 ) were downloaded from UNIPROT [28] , and each was used as a seed sequence for the SAM-T2K iterative alignment-building algorithm [23] . For BRCA1 , only amino acid residues in the BRCT domains ( 1649–1859 ) were aligned . We used the SAM w0 . 5 program to apply sequence weighting and regularization with Dirichlet mixtures [29] to each resulting alignment and to produce a profile hidden Markov model [30] . The TP53 and BRCA1 alignments and hidden Markov models are available upon request . We trained a soft margin Support Vector Machine classifier with a radial basis kernel using the e1071 package in R [31] . The Support Vector Machine algorithm optimizes a vector of weights ( one weight for each training example ) and a bias parameter b . The parameters g ( radial basis kernel width ) and C ( penalty for violating the soft margin ) were optimized on the training set with grid search using default parameters . Each of the 28 missense variants was then scored with the discriminant function where l is the number of examples in the training set , yi is the class label of each example in the training set ( for deleterious/nonfunctional variants yi = −1 and for neutral/functional variants yi = 1 ) , and is the value of the radial basis kernel function given and training example . Variants are classified as deleterious/nonfunctional if and neutral/functional if . The Naïve Bayes algorithm estimates the probability that each variant belongs to deleterious or neutral classes by applying the Bayes rule: where the prior class probability P ( C ) is the fraction of deleterious ( or neutral ) missense variants in the training set and each feature Xi is assumed to be conditionally independent of the k − 1 other features , given its class membership , so that where P ( Xi | C ) is estimated from the training set . We used the Naïve Bayes method in R's e1071 package . Each feature was approximated to be normally distributed and no smoothing was applied to the feature distributions . We used the rpart package in R [32] to train a Decision Tree with the following parameters: minsplit = 20 ( minimum number of observations required at a tree node before a split is attempted ) and cp = 0 ( no pruning of tree regardless of whether a split will improve model fit ) . To reduce overfitting , we pruned the resulting tree using the standard heuristic “1 Standard Error rule” [33] and 10-fold cross-validation . According to the 1 Standard Error rule , the pruned tree with best generalization properties has a cross-validation error on the training set 1 Standard Error worse than the tree with the lowest cross-validation error . The pruning process yielded a reduced set of features: Φ and Ψ mainchain dihedral angles , normalized solvent accessibility of wild-type , Grantham difference , volume change , relative entropy , and positional hidden Markov model conservation score . We used the randomForest package in R [34] to train a Random Forest , an algorithm based on a majority vote of a large number of decision trees , in which the candidate features at each tree node are randomly sampled [17] . The user-defined input parameters to randomForest are total number of trees in the forest and mtry ( number of randomly sampled features considered as candidates for a split at each tree node ) . Both were selected with grid-search optimization as described for the Support Vector Machine [31] . Predictions of Naïve Bayes , Decision Tree , and Random Forest are in the form of class conditional probabilities , where the two classes are D ( deleterious/nonfunctional ) and N ( neutral/functional ) . For each example , the classifiers report P ( D | ) ( probability that the variant is deleterious , given feature vector ) and P ( N | ) ( probability that the variant is neutral , given feature vector ) . To evaluate accuracy , true positive rate , true negative rate , and Matthews correlation coefficient , we classified variants as deleterious if P ( D | ) > 0 . 5 and neutral otherwise . To compute ROC curves , we used the log likelihood ratio as the output score of Naïve Bayes , Decision Tree , and Random Forest . The SIFT algorithm [19] predicts the probability that a missense mutation occurs at a given alignment position . Variants that occur at conserved alignment positions are expected to be tolerated less than those that occur at diverse positions . The algorithm uses a modified version of PSI-BLAST [35] and Dirichlet mixture regularization [29] to construct a multiple sequence alignment of proteins that can be globally aligned to the query sequence and belong to the same clade . We used the SIFT server ( http://blocks . fhcrc . org/sift/SIFT . html ) , with PSI-BLAST search set to the Swissprot-TrEMBL protein sequence database [36] . Both the full-length human BRCA1 sequence ( amino acid residues 1–1863 ) and the BRCT C-terminal domain sequence only ( amino acid residues 1649–1859 ) were submitted to the server . For the full-length sequence , SIFT reported low confidence predictions for 34 out of 36 missense variants . Consequently , we based our SIFT predictions on the C-terminal domain sequence . To compute accuracy , true positive rate , true negative rate , and Matthews correlation coefficient , we used the binary class predictions of the SIFT server ( deleterious or neutral ) , based on the default SIFT threshold ( tolerated mutation probability > 0 . 05 ) . For ROC analysis , we used the raw SIFT probabilities , which range from 0 to 1 . The Ancestral Sequence classifications were computed as described [9 , 11] . Each position in an alignment of eight mammalian BRCA1 orthologs identified as giving best results by Pavlicek et al . was categorized as fixed ( completely conserved ) , conserved ( substitution of similar amino acid residues ) , or nonconserved ( dissimilar amino acid residues or gaps ) . Any substitution at a fixed position and any nonconservative substitution at a conserved position is classified as deleterious . Amino acid residue similarity is based on the Gonnet PAM250 score ( i . e . , the likelihood that amino acid residue A has mutated into amino acid residue B in a pair of sequences that have diverged by 250 mutations per 100 amino acid residues of sequence ) [37] . The Align-GVGD method calculates two scores for each amino acid residue substitution , Grantham Deviation ( GD ) and Grantham Variation ( GV ) , based on a modified Grantham distance measure [18 , 38] . The scores define four categories of missense variants: “Enriched deleterious 1” variants occur at invariant alignment positions for which the substitution is outside the range of variation observed at the position ( GV = 0 , GD > 0 ) ; “Enriched deleterious 2” occur at variable alignment positions containing physiochemically similar amino acid residues where the substitution is outside the range of observed variation ( 0 < GV < 61 . 3 , GD = 0 ) ; “Enriched neutral 1” occur at variable positions containing physiochemically similar amino acid residues where the substitution is inside the range of variation ( GV > 0 , GD = 0 ) ; and “Enriched neutral 2” occur at variable positions containing dissimilar amino acid residues where the substitution is slightly outside the range of variation ( GV > 61 . 3 , 0 < GD < 61 . 3 ) . We classified variants using first an alignment of placental mammals , a marsupial ( gray short-tailed opposum ) , chicken , frog , and the pufferfish Tetraodon ( “Align-GVGD Tnig” ) , and second an alignment that also includes the sea urchin Strongylocentrotus purpuratus ( “Align-GVGD Spur” ) . Accuracy , true positive rate , true negative rate , and Matthews correlation coefficient were evaluated by reducing the four categories to deleterious/nonfunctional or neutral/functional . Variants may have GV and GD values that do not match any of the four categories ( e . g . , a variant with GV = 80 and GD = 80 ) , which lowers coverage , Matthews correlation coefficient , true positive , and false positive rates . For ROC analysis , rather than fixing thresholds on GV and GD at 61 . 3 for each substitution , we considered the number of true positives and false positives over a range of thresholds , from the smallest to largest values of GV and GD in our dataset ( 0 to 215 ) . A Rule-based decision tree is a classification tree with human-designed rules that uses both structure- and sequence-based information , implemented in PERL [7] . Rule-based Decision Tree classifies a missense variant as either deleterious/nonfunctional or neutral/functional , but does not compute numerical scores . The structural models of all BRCA1 BRCT missense variants were visually compared with the wild-type structure ( 1t29 ) using the molecular graphics program Chimera [39] . We explored changes in hydrogen bonding patterns and geometric properties of the molecular surface with Chimera's FindHBond and MSMS routines . To visualize the distribution of amino acid residue conservation on the protein surface , the RenderByAttribute routine was used , with coloration defined by percent conserved in a hand-edited SAM-T2K alignment of BRCA1 orthologs . Species used in this alignment were Homo sapiens ( AAA 73985 ) , Pan troglodytes ( AAG43492 ) , Gorilla gorilla ( AAT44835 ) , Pongo pygmaeus ( AAT44834 ) , Macaca mullata ( AAT44833 ) , Canis familiaris ( AAC48663 ) , Bos taurus ( AAL76094 ) , Monodelphis domestica ( AAX92675 ) , Mus musculus ( AAD00168 ) , Rattus norvegicus ( AAC36493 ) , Gallus gallus ( AAK83825 ) , Xenopus laevis ( AAL13037 ) , and Tetraodon nigroviridis ( AAR89523 ) . A highly conserved surface patch was identified as a possible binding site and subjected to further analysis . We used DELPHI [40] to compute the electrostatic surface potential at the putative binding site for the wild-type structure and for models of two solvent-exposed variants characterized as deleterious in our functional assays ( T1685A and R1753T ) [14] . The solvent relative dielectric constant was set to 4 . 0 , the protein relative dielectric constant to 20 . 0 , and ionic strength to the physiological value of 0 . 2 mM . Charges were estimated with the united atom AMBER model [41] . The proteins were prepared for DELPHI by adding heavy atoms missing from the 1t29 crystal structure with MODELLER's COMPLETE_PDB routine and adding hydrogens with REDUCE [42] , then visualized in Chimera with a GRASP surface representation [43] .
Based on ROC analysis , the supervised learners Random Forest , Support Vector Machine , and Naïve Bayes yield the most reliable computational classifications of the 36 variants ( Figure 2 ) . The area under the ROC curve ( AUC ) quantifies the probability that a classifier will give a randomly drawn deleterious example a lower score than a randomly drawn neutral example . AUC is 0 . 992 for Random Forest , 0 . 947 for Support Vector Machine and Naïve Bayes , 0 . 86 for Align-GVGD Tnig , 0 . 852 for Align-GVGD Spur , 0 . 783 for SIFT , and 0 . 738 for Decision Tree ( Figure 2 ) . The Decision Tree algorithm appears to overfit the training set and generalizes less well than the other supervised learners . Three of the supervised learning algorithms ( Naïve Bayes , Support Vector Machine , and Random Forest ) produce the best classifications of the 36 variants , as measured by accuracy , true positive rate , true negative rate , Matthews correlation coefficient , and coverage , using default thresholds ( Table 1 ) . According to these statistical measures , the best sequence analysis methods are Ancestral Sequence and Align-GVGD . Random Forest , Naïve Bayes , and Support Vector Machine are the most accurate scoring predictors , according to the AUC . The methods rankings are slightly different when evaluated by threshold-dependent statistics that reduce predictive scores to deleterious/neutral classes or by the score-based and threshold-independent ROC statistic of AUC . We applied the top performing algorithms ( Naïve Bayes , Support Vector Machine , and Random Forest ) to predict the impact of 54 BRCA1 UCVs listed in BIC that ( a ) are located in the BRCT domains , and ( b ) have not been functionally characterized by our transactivation assays . Based on a majority vote of the computational predictors , we computed a “consensus prediction” for each UCV ( Figure 3 ) . We provide structural explanations for the impact of as many variants as possible , and indicate where the predictions are supported by biochemical experiments found in the literature ( Table S2 ) . The predicted deleterious UCVs are predominantly in the core secondary structure elements of the BRCT domains , rather than in loops , particularly in the β sheet of BRCT-N ( β1 , β3 , and β4 ) , and in helix α′3 and the turn connecting helix α′1 and strand β′2 in BRCT-C ( Figure 4 ) . We observed a patch of highly conserved amino acid residues that form a groove on the BRCA1 surface , on the opposite face from the known phosphopeptide binding cleft ( Figure 5A–5C ) . These residues are T1684 , T1685 , H1686 , K1711 , W1712 , and R1753 . Both T1685 and H1686 have been shown to be highly sensitive to mutation , and our companion paper [14] contains new experimental evidence that R1753T has markedly reduced transactivation activity in both yeast and mammalian cells . The groove residues form hydrogen bonds with each other and several other conserved residues , including S1651 , V1687 , T1681 , G1706 , E1731 , E1735 , and P1749 , producing two hydrogen bonding networks . The first network is found in BRCT-N ( S1651 , T1684 , T1685 , H1686 , V1687 , G1706 , K1711 , W1712 , E1731 ) and the second network connects BRCT-N residues with the linker region that connects BRCT-N and BRCT-C ( E1735 , P1749 , R1753 ) ( Figure 5A–5C ) . All the residues lining this groove are completely conserved in our alignment of BRCA1 orthologs , except for T1684 , which is conserved in all orthologs except for Tetraodon ( pufferfish ) , the organism most distant from human in our alignment ( Figure 5A ) . Previous studies have shown that G1706 and P1749 are also sensitive to mutation [44 , 45] ( S . Marsillac , 2006 , private communication ) . The proposed binding site would be specific to BRCA1 , as most of these positions ( except for H1686 ) are not highly conserved across tandem BRCT repeats in MDC1 , PTIP , BARD1 , and 53BP1 . The solvent-exposed missense variant R1753T found at the proposed binding site has <20% of the wild-type transactivation activity in yeast and <5% in mammalian cells [14] , suggesting that the wild-type arginine amino acid residue might be important for binding of BRCA1 to a protein partner ( or nucleic acid ligand ) . Although the mechanism of the BRCA1 BRCT domains in transactivation is not known , it is believed to depend on interactions with a variety of partners [46 , 47] . The mutation of R1753 to a threonine changes the local electrostatic surface potential from primarily positive and neutral ( depicted as blue and white ) to negative ( red ) ( Figure 5D ) . This change may weaken the binding of protein partner ( s ) or nucleic acid ligand ( s ) necessary for transactivation .
Several studies of the BRCA1 BRCT domains have suggested that there may be surface patches that interact with protein partners [60–63] . In previous work , we predicted that a groove formed by both BRCT repeats ( near nonfunctional variants L1657 and K1702 ) and the ridge that delimits the groove ( near nonfunctional variant E1660 ) constitutes such a surface patch [7] . Our prediction was subsequently confirmed through X-ray crystallographic studies , which revealed a site where the phosphorylated peptides of BACH1 and CtIP have been found to bind [64 , 65] . Importantly , several missense variants were found to disrupt this interaction [64 , 65] , suggesting that clustering of deleterious variants at solvent-exposed amino acid residue positions is indeed a useful indicator of binding site location . There is a large literature on the general topic of the relationship between deleterious mutations and binding sites [66–74] . Two surface variants found to be deleterious to BRCA1 transactivation activity in our companion paper ( R1753T and T1685I ) [14] lie on a highly conserved patch of amino acid residues , forming an exposed groove . The R1753T variant yields a changed electrostatic surface potential , which may be sufficient to disrupt the binding of BRCA1 to a protein partner or nucleic acid ligand important for transactivation . Following the logic that predicted the BACH1/ CtIP binding site , we suggest that this groove may be a previously uncharacterized binding site , whose disruption inactivates BRCA1 transactivation function . Accordingly , we are currently testing the binding of several candidate protein partners to the predicted binding site using site-directed mutagenesis and a yeast two-hybrid assay . Biochemical assays can play an important role in identifying deleterious UCVs in cancer susceptibility genes [75] , but the work is labor-intensive and time-consuming . We estimate that , on average , an assay for one BRCA1 UCV in a mammalian cell system costs US$125–US$150 and requires three weeks of personnel time , from ordering primers to final results . The time can be reduced by processing the variants in batches . To assay the 54 uncharacterized BRCT BRCA1 missense variants found in the BIC database ( April 2006 ) would take approximately 18 months of personnel time . Accurate computational classification of UCVs can significantly reduce the required time by prioritizing UCVs most likely to be deleterious . Importantly , while computational classification and functional assays can contribute to medical decision making , other factors such as family history , co-occurrence with known deleterious mutations , and studies of patient tumor tissue will continue to be important in a clinical setting . We have applied supervised learners trained on the TP53 variant set to prediction of UCVs in BRCA2 , with promising results ( unpublished data ) . We are currently exploring whether this training set and our current set of features can be used to evaluate UCVs in other genes associated with familial cancer syndromes: MLH1 , MSH2 , MSH6 ( hereditary nonpolyposis colon cancer ) , APC ( familial adenomatous polyposis ) , MYH ( MYH adenomatous polyposis ) , and P16 ( melanoma ) . We have applied a modified version of this method to classify all human amino-acid changing SNPs found in the dbSNP database [76] as deleterious or neutral [56] . The SNPs were classified with a support vector machine trained on amino acid residue substitutions from more than 1 , 500 human proteins . Because X-ray crystal structures are not available for most human proteins [77] , we built homology models with an automated modeling pipeline MODPIPE that relies on the MODELLER package for fold assignment , sequence-structure alignment , model building , and model assessment [78] . A small number of these predictions have been validated by biochemical and epidemiological studies found in the literature . We are exploring the extent to which a decision rule learned with a training set of variants from one protein , such as TP53 , can be generalized to variants from other proteins . One possibility is that most deleterious missense mutants do not affect specific binding interactions , but are instead slightly destabilizing [79] . If this is true , a training set of missense variants from a protein with similar stability to the protein of interest may be the best choice . Other possiblities include training on a protein sharing GO terms , or from the same fold family ( all-alpha , all-beta , alpha-beta , etc . ) as the protein of interest . We are working on generating large variant datasets from selected proteins to test these hypotheses . In summary , we have systematically and comprehensively evaluated structure- and sequence-based computational prediction methods applied to variants in the BRCA1 BRCT domains and developed detailed structural explanations for the measured and predicted impact of 49 BRCA1 variants . When combined with 16 carefully selected predictive features , the best-supervised learning algorithms are in greater agreement with experimental results than has been reported previously . The increased use of sequencing methods to genotype individuals at risk for inherited cancers and the observation that sequence variation is greater in ethnic minorities than in Caucasians highlight the need for improved methods of UCV risk assessment . Bioinformatics approaches including supervised learning algorithms , protein structure modeling , and evolutionary sequence analysis can contribute to an integrated approach to risk assessment by increasing coverage of classified UCVs more rapidly than is possible by functional assays . In the future , when clinicians counsel patients about their cancer risk , they will be able to take advantage of these bioinformatics prediction methods . Finally , successful generalization of these methods to a large number of disease-associated genes will play an important role in reducing the growing number of loci , variants , and phenotypes that confound modern whole genome disease-association studies .
Accession numbers from the National Center for Biotechnology Information ( http://www . ncbi . nlm . nih . gov ) are shown in Table 3 . | A significant number of breast and ovarian cancers are due to inherited mutations in the BRCA1 and BRCA2 genes . Many women who receive genetic testing for these mutations are found to have variants of the genes that result in changed amino acids in the BRCA1 or BRCA2 proteins . The effect of these variants on cancer risk is not well-understood , posing a problem for patients and their health providers . We describe computational biology methods that predict and analyze the impact of 36 BRCA1 variants on protein function . The predictions are validated by biochemical assays of BRCA1 in yeast and mammalian cell cultures . The speed and accuracy of the computational methods is well-suited to rapid evaluation of large numbers of variants in genes that predispose to inherited diseases . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Supporting",
"Information"
] | [
"oncology",
"in",
"vitro",
"homo",
"(human)",
"computational",
"biology"
] | 2007 | Functional Impact of Missense Variants in BRCA1 Predicted by Supervised Learning |
The characterization of protein interactions is essential for understanding biological systems . While genome-scale methods are available for identifying interacting proteins , they do not pinpoint the interacting motifs ( e . g . , a domain , sequence segments , a binding site , or a set of residues ) . Here , we develop and apply a method for delineating the interacting motifs of hub proteins ( i . e . , highly connected proteins ) . The method relies on the observation that proteins with common interaction partners tend to interact with these partners through a common interacting motif . The sole input for the method are binary protein interactions; neither sequence nor structure information is needed . The approach is evaluated by comparing the inferred interacting motifs with domain families defined for 368 proteins in the Structural Classification of Proteins ( SCOP ) . The positive predictive value of the method for detecting proteins with common SCOP families is 75% at sensitivity of 10% . Most of the inferred interacting motifs were significantly associated with sequence patterns , which could be responsible for the common interactions . We find that yeast hubs with multiple interacting motifs are more likely to be essential than hubs with one or two interacting motifs , thus rationalizing the previously observed correlation between essentiality and the number of interacting partners of a protein . We also find that yeast hubs with multiple interacting motifs evolve slower than the average protein , contrary to the hubs with one or two interacting motifs . The proposed method will help us discover unknown interacting motifs and provide biological insights about protein hubs and their roles in interaction networks .
Protein–protein interactions play a central role in many cellular processes , ranging from signal transduction to formation of cellular macrostructures and cell cycle control [1–3] . Recently , several techniques such as two-hybrid assays [4–6] and affinity purifications followed by mass spectrometry [7–9] have enabled large-scale identification of protein–protein interactions . While these efforts provide rich lists of interacting proteins , they do not produce information about the specific interfaces involved in each interaction . Proteins interact through a limited set of interface types [3 , 10 , 11] . These interfaces are usually key determinants of the function . Therefore , narrowing down protein–protein interactions to interactions between specific protein components ( e . g . , a domain , sequence segments , a binding site , or a set of residues ) is important for a more accurate characterization of the function of proteins and their complexes . Identifying the protein interfaces that mediate interactions may also be useful for the prediction of unknown protein–protein interactions [12 , 13] , for homology-based protein annotation methods [14] , and for relating gene essentiality and network topology [15] . Traditionally , the description of protein interactions in terms of the interacting components has been based on protein structural domains [16] , protein functional sites [17] , and protein patches [18] . However , fully characterizing protein surfaces that are in contact with each other during an interaction requires the determination of the structure of protein complexes by X-ray crystallography or NMR spectroscopy . These methods are not always applicable and thus the number of known 3-D atomic structures of proteins and their complexes is limited . As a result , accurate and general computational methods for identifying motifs involved in protein–protein interactions are needed . Recently , several methods [19–25] have been developed to describe protein–protein interactions in terms of interacting protein domains , as defined in the Structural Classification of Proteins ( SCOP ) [26] , PFAM [27] , and InterPro [28] databases . However , while these methods find interactions between predefined protein domains , interactions between undefined domains remain undetected . Sequence-based methods overcome this problem by identifying sequence signatures that consistently co-occur in pairs of interacting protein sequences [29] , while structure-based methods can predict the amino acid residues that are in contact during a protein–protein interaction , but require information about the structures of both proteins [30–33] . Recently , Kim et al . used known protein interactions and structures to characterize the interfaces between two interacting proteins [15] . They found that some previously accepted relationships between network topology and genomic features [34–36] are actually more reflective of the number of distinct binding interfaces . For example , highly connected proteins in the network ( i . e . , hubs ) with multiple interfaces are twice as likely to be essential as hubs with one or two interfaces . The findings of Kim and coworkers clarify some previous analyses that related the observed essentiality of hubs with their high number of interacting partners [34 , 37] or with their interactions to other hubs [38] . Kim et al . also demonstrated that the evolutionary rate is significantly lower for multi-interface hubs than for the average protein , but not so for hubs with one or two interfaces . Here , our basic assumption is that proteins with overlapping sets of interacting partners tend to interact with the common partners through the same interacting motif , such as a domain , sequence segments , a binding site , or a set of residues . A similar assumption has been previously used to annotate protein sequences [14 , 39–41] . We first tested this assumption based on databases of protein interactions [42] and protein domains defined in SCOP [26] , observing that the assumption holds true for highly connected proteins ( i . e . , hubs in a protein–protein interaction network ) . Building on this validation , we then developed a method for identifying interacting motifs ( iMotifs ) , which has been implemented within the protein–protein interaction framework and integration engine PIANA ( Protein Interactions and Network Analysis ) [42] . iMotifs are not required to be of any particular structural type or size , thus allowing us to characterize proteins and their interactions at different levels of resolution , ranging from full proteins to small binding sites . In contrast to other methods , our approach is not limited to finding predefined classes of interacting motifs , such as SCOP domains or PROSITE functional sites , and can be used to identify unknown interacting motifs . Moreover , the sole input for our method is binary protein interactions; neither structure nor sequence information is required to assign iMotifs to proteins . Two main objectives have been addressed in this work . The first objective was to demonstrate whether protein interactions alone can be used to infer interacting motifs . The positive predictive value of our method in detecting proteins with common SCOP families was 75% at sensitivity of 10% , and the Spearman correlation coefficient between the number of iMotifs assigned to proteins and the number of interfaces found by Kim et al . [15] was 0 . 57 . The second objective was to examine if the conclusions on protein hubs of Kim et al . [15] hold for our iMotifs assignments . The results demonstrate that protein hubs with multiple iMotifs are more likely to be essential than hubs with one or two iMotifs and that protein hubs with multiple iMotifs evolve slower than the average protein in the dataset , as opposed to hubs with one or two iMotifs .
The basic assumption behind this work is that proteins with overlapping sets of interaction partners tend to interact with those partners through a common interacting motif . The validity of this assumption was tested on a nonredundant set of 368 proteins with known SCOP domains ( Material and Methods ) . Although SCOP does not classify proteins by their interfaces , SCOP domains were used as surrogates for iMotifs because protein interaction types can be defined by the domains in the interacting proteins [43] . We found the number of common interaction partners ( N ) to be a good indicator of the probability of two proteins having a domain in the same SCOP family , especially for highly connected proteins ( Figure S1 ) . For example , 73% of protein pairs with 50–60 common interaction partners shared a SCOP domain . We also studied other metrics to measure the similarity between two sets of interaction partners , but none of them outperformed N at the identification of protein pairs with a common domain family ( Figures S1 and S2A , and Table S1 ) . It is worth noting that our assumption relies on the binary nature of the input interactions . Two proteins tend to have a common interacting motif only if they share direct physical interactions with the same partner ( s ) . However , the likelihood of two proteins sharing a SCOP domain was lower by solely using yeast two-hybrid experimental data , a detection method that is more likely to contain binary protein interactions than other experimental methods [44] ( Figure S2B ) . Based on the observation that highly connected proteins with common interaction partners tend to interact with them through a common interacting motif , we have developed a method that groups proteins with similar interacting motifs ( Figure 1 and Material and Methods section ) . Briefly , the procedure is carried out in four steps: 1 ) build a protein–protein interaction network; 2 ) initialize a cluster interaction network by assigning each protein of the network to a cluster; 3 ) iteratively create new clusters by fusing similar clusters ( allowing a protein to be in more than one cluster ) until the similarity score drops below a predefined threshold; and 4 ) label with a different interacting motif identifier ( iMotif ) each cluster with more than one protein and derive iMotif–iMotif interactions from the clustered network . In step 3 ) , the similarity score between two clusters is their number of common interacting partners in the cluster interaction network ( N ) . Assigning an iMotif to a group of proteins simply establishes that they have a certain feature that allows them to interact with the same set of partners , without determining the size , sequence , or structure of that feature ( Figure 2A ) . Thus , an iMotif can be an interface consisting of a set of domains or only a specific constellation of a small number of residues ( Figure 2B ) . The definition of iMotifs depends on a similarity metric and its threshold . Thus , different thresholds or metrics produce different iMotifs , corresponding to different levels of resolution in the description of protein–protein interactions . For example , the method can be applied at the resolution of domains from SCOP [26] , and PFAM [27] , or at the higher resolution of functional sites from PROSITE [45] . In this section , we have evaluated the method on a nonredundant set of proteins ( Material and Methods ) for three different tasks: ( i ) detecting proteins with common SCOP domain families; ( ii ) predicting SCOP domain–domain interactions observed in the Protein Data Bank ( PDB ) [46]; and ( iii ) predicting the number of distinct binding interfaces as defined by Kim et al . [15] . Therefore , in the evaluation , iMotifs effectively represent SCOP family domains ( for the first two tasks ) and structural binding interfaces ( for the third task ) . We evaluated the ability of the method to detect proteins with a domain in the same SCOP family ( Methods ) . Using an N threshold of 30 common interaction partners , our method achieves a positive predictive value of ∼75% , sensitivity of ∼10% , and applicability of ∼20% ( Figure 3 ) . The positive predictive value drops to ∼50% for N of 15 , indicating that the accuracy of our method in detecting proteins with common SCOP family domains proportionally decreases with the number of common interacting partners . Therefore , the method should be preferentially applied to assigning interacting motifs to highly connected proteins . The growth of the interactome data [47 , 48] is likely to make the approach more applicable in the future . Nevertheless , the applicability can already be increased at the expense of the positive predictive value by using other similarity metrics ( Table S1 ) . We provide a complete list of iMotif assignments for the test set ( Table S2 ) . Domain–domain interactions can be predicted from the iMotif–iMotif interactions found by the method ( Materials and Methods ) . We evaluated the accuracy of these predictions with respect to domain interactions in the PDB . Our method achieves a positive predictive value of ∼65% for ∼5% of the proteins in the test set ( Figure S4 ) , suggesting that the method can be applied to the prediction of domain–domain interactions when a sufficiently large and varied sample of protein interactions is known . However , with the available interaction data , other methods that rely on both interaction networks and predefined domains [19–22] may be better suited than our approach for predicting domain–domain interactions . Kim et al . used protein 3-D structures and binary protein interactions to make inferences about the number of binding interfaces of proteins [15] . We tested whether there is a correlation between the number of binding interfaces found in their work and the number of iMotifs predicted by our method ( Figure 4 ) . The number of protein interfaces indeed correlates with the number of predicted iMotifs per protein ( e . g . , for N of 20 , rs is 0 . 57 and p-value 0 . 01 ) . The number of iMotifs assigned to proteins by our method tends to be higher than the number of binding sites defined by Kim et al . This might be attributed to two factors: ( i ) current structural data do not contain all possible protein–protein interactions , resulting in an underestimation of the number of binding sites assigned by the method in [15] , and ( ii ) the lack of coverage of the interactome space , which results in an overestimation of the number of iMotifs per protein assigned by our method . The second factor is addressed by using sequence information to merge similar iMotifs ( below ) . Using an N threshold of 20 , our method assigned 12 , 342 iMotifs to 2 , 014 of the 5 , 571 hub proteins in PIANA ( i . e . , proteins with 20 or more interaction partners ) , resulting on average in 8 . 6 iMotifs per hub . The percentage of hubs with one or two iMotifs was 46% ( 241 hubs had one iMotif; 689 hubs had two iMotifs ) . We studied the correlation between the number of iMotifs assigned to a hub and its number of interactions , finding no relationship between the two variables ( Spearman correlation coefficient is −0 . 002 with p-value 0 . 94 ) . A complete list of iMotif assignments for all hub proteins in PIANA is in Table S3 and the number of iMotifs per hub is in Table S4 . Similarly to Kim and co-workers' results , [15] , we found that yeast hubs with multiple iMotifs are more likely to be essential than those with one or two iMotifs ( singlish-iMotif ) ( Table 1 ) . Furthermore , we observed a correlation ( rs is 0 . 61 and p-value is 1 . 64 × 10−5 ) between the number of iMotifs in yeast hubs and the fraction of essential proteins ( Figure 5A ) . We compared the correlation between iMotifs and essentiality to the correlation between the number of interactions of hubs and essentiality to confirm that the first was not a direct consequence of the second ( Figure 5B ) . These results suggest that the number of iMotifs predicted for a protein could be used for selecting biologically relevant candidates for gene deletion experiments . A common measure of evolutionary rate is the dN/dS ratio ( the ratio of nonsynonymous to synonymous substitutions ) [49] . Kim et al . found that multi-interface hubs have a lower evolutionary rate than the average protein in their data , but the same was not true for singlish-interface hubs . Our results are in agreement with their findings . Multi-iMotif hubs , in contrast to singlish-iMotif , evolve significantly slower than the average protein in our dataset ( Table 2 ) . However , the evolutionary rate difference between multi- and singlish-iMotif hubs ( i . e . , 0 . 062 and 0 . 056 , respectively ) was not found to be significant ( p-value of 0 . 21 ) . Sequence patterns for each iMotif were generated using the PRATT program [50] ( Methods ) . Briefly , PRATT identifies sequence patterns common to a set of sequences . In this work , we selected significant sequence patterns for each iMotif by maximizing the number of proteins within the iMotif that contained the pattern . The significance ( i . e . , p-value ) of a sequence pattern assigned to an iMotif depends on the occurrence of the pattern in the iMotif with respect the whole dataset . As shown on Figure 6 , 80% of iMotifs had a specific sequence pattern contained in at least 74% of their proteins ( using a p-value cutoff of 10−8 ) . A list with the best sequence pattern for each iMotif is provided in Table S5 . Interestingly , a similar analysis based on Pfams assignments to iMotifs showed a different trend ( i . e . , very few iMotifs had most of their proteins described by a Pfam ) . For example , as shown on Figure S6 , only 10% of all iMotifs had a specific Pfam in at least 28% of their proteins ( p-value cutoff of 10−8 ) . Such a difference can be explained by the fact that many interactions are carried out by short sequence patches [3 , 51] , while Pfam families usually consist of long structured protein regions . As indicated above , incompleteness in interaction data may result in artificially high numbers of iMotifs . This overestimation can be reduced by merging iMotifs with a common sequence pattern ( Material and Methods ) . Fusing iMotifs based on sequence pattern similarity decreased the average number of iMotifs per hub from 8 . 6 to 4 . 2 . This reduction , in turn , increased the correlation between the number of binding sites from Kim et al . [15] and the number of iMotifs in the test set proteins ( Spearman correlation coefficient was 0 . 59 with p-value of 0 . 001 ) .
We described , implemented , and evaluated a method that relies solely on binary protein interactions to identify interacting motifs ( iMotifs ) and their interactions . Our approach obtained high positive predictive value for identifying proteins with domains from the same SCOP family and predicting domain–domain interactions . We also analyzed hub proteins and their properties based on the number of iMotif assigned to them , obtaining similar findings to those in an independent approach that rely on protein structure information [15] . Recent estimates suggested that only one-fifth of interaction types are known [43] . Therefore , current knowledge of protein structures is not sufficient to describe all protein interaction types . Our approach , in contrast to other previously described methods , accomplishes three different objectives: ( i ) it predicts the number of different iMotifs in a protein , ( ii ) it classifies proteins by their predicted iMotifs , and ( iii ) it predicts interactions between the iMotifs . The method can identify iMotifs independently of structural or sequence information; it can assign an iMotif to two structural domains or two iMotifs to a single domain . Since the resolution at which iMotifs describe protein interfaces depends on the similarity metric used and the threshold applied by the method , iMotif assignments can be used to infer whether the interaction is mediated through multiple , single , or partial domains . On the one hand , setting a high threshold on the number of common interaction partners ( N ) will assign few iMotifs to reduced sets of proteins ( i . e . , very specific and restrictive iMotifs ) . On the other hand , using low N thresholds will assign the same iMotif to broad numbers of proteins ( i . e . , very unspecific and general iMotifs ) . We showed that the method works better for highly connected proteins and using high values for N . Moreover , our approach is not limited to finding predefined classes of protein components and thus allows us to predict new types of interacting motifs . For example , an iMotif can be mapped to a predefined class ( e . g . , a SCOP domain or a PROSITE functional site ) by examining the known classes assigned to proteins with that iMotif . Therefore , iMotifs that remain unmapped are likely candidates for unknown classes . Such predictions may prove useful for target selection in structural genomics . Relying solely on experimentally detected interactions affects the accuracy of our method . It has been shown that high-throughput experiments have limited reliability and that many of the detected interactions are probably not direct ( i . e . , they are carried out through a third protein ) or do not even exist ( i . e . , false positives ) [52] . However , we did not observe an improvement by solely using interactions from yeast two-hybrid assays ( Figure S2B ) , the high-throughput method that is best suited to detect direct interactions . As more interaction data becomes available , we will reexamine the effects of restricting the method to employ interactions specifically labeled as “direct” [6] . One way of avoiding these limitations is to calculate similarity scores using families of proteins instead of absolute numbers of protein partners . This will prevent assigning the same iMotif to proteins that have many common partners but all of them belong to a single protein family . Removal of redundancy from the sets of partners indeed increases the percentage of identified protein pairs with a common domain family ( Figure S7 ) . The combination of iMotif assignments with sequence search methods identified specific sequence patterns in iMotifs . We found that most iMotifs had a significant sequence pattern that was contained in most of the iMotif proteins . These patterns , which could be responsible for the iMotif proteins common interactions , could then be used to: ( i ) localize the iMotif in the protein sequence , ( ii ) assign iMotifs to proteins for which no interaction data is yet available , and ( iii ) predict interactions between proteins that contain patterns assigned to two interacting iMotifs . Our iMotif assignments are similar to those obtained using an independent approach , which relies not only on known protein–protein interactions , but also on protein structure information [15] . In agreement with the results of Kim et al . , we observe different properties between hubs with multiple iMotifs ( multi-iMotif ) and hubs with one or two iMotifs ( singlish-iMotif ) . In particular , we find that ( i ) multi-iMotif hubs are more likely than singlish-iMotif hubs to be essential for cell viability , and ( ii ) multi-iMotif hubs , in contrast to singlish-iMotif hubs , evolve slower than the average protein . Furthermore , we have also observed a correlation between the number of iMotifs of a hub and its essentiality for cell survival . The properties observed for hubs with respect to their number of iMotifs may reflect the difference between proteins with multiple simultaneously possible interactions ( multi-iMotif hubs are probably involved in permanent complexes ) and proteins with multiple exclusive interactions ( for singlish-iMotif hubs involved in transient interactions ) . This is in agreement with the previous observation that interfaces of transient protein–protein interactions are less restricted in evolution than interfaces in permanent complexes [53] . Our results extend the findings and conclusions of Kim and co-workers [15] to proteins of unknown structure . Thus , inferring interacting motifs from protein interactions is likely to be helpful for providing biological insights about hubs for which no structural information is available .
Protein–protein interactions from DIP 2006 . 01 . 16 [54] , MIPS 2006 . 01 [55] , HPRD 2005 . 09 . 13 [56] , BIND 2006 . 01 [57] , and two recent high-throughput experiments [5 , 6] were integrated using PIANA version 1 . 2 [42] , allowing us to work with a large set of 363 , 571 interactions between 42 , 040 proteins . PIANA represents protein interactions as a network where the nodes are proteins and the edges are interactions between the proteins . In such a network , a set of proteins linked to protein pj ( i . e . , physically interacting with pj ) is named “partners of pj” . In such a network , we define hubs as proteins with 20 or more partners . The average number of interactions per hub in our dataset was ∼49 . PIANA builds the protein interaction network by retrieving partners for an initial set of proteins . To avoid a positive bias in the method evaluation , interactions inferred from 3-D structures were not used in this work . Protein domain assignments and classification were obtained from the SCOP release 1 . 69 [26] . Here , domains are defined at the SCOP family level . Thus , domain–domain interactions refer to SCOP family interactions . The number of protein binding interfaces for hub proteins was obtained from the Structural Interaction Network 2 . 0 [15] . A list of ORFs essential for the survival of the yeast cell was obtained from the Saccharomyces Genome Deletion Project [58] . The evolutionary rates ( dN/dS ) of yeast proteins were taken from the adjusted values given by Wall et al . [49] . The assignment of iMotifs to a set of proteins is carried out in a four-step procedure ( Figure 1 ) : First , build the protein interaction network . Second , initialize a cluster interaction network ( i . e . , nodes are clusters that contain one or more proteins , and edges are interactions between clusters ) by assigning each protein of the protein interaction network to a different cluster . Each cluster ( containing one protein pj ) interacts with those clusters that contain a partner of pj in the protein interaction network . Third , iteratively create new clusters by fusing the most similar clusters until the similarity score drops below a predefined threshold . Two clusters are similar if they share a minimum number of common interacting partners ( N ) . Thus , the similarity score between two clusters is their number of common partners in the cluster interaction network . Other similarity metrics were considered , but none outperformed the use of N ( Figure S1 ) . When fusing two clusters , the resulting cluster inherits interactions that were common to both fused clusters . Since proteins may have multiple interfaces , all initial clusters ( from step 2 ) remain in the cluster interaction network even after being fused to another cluster . Interactions between non-initial clusters are not considered for calculating the similarity scores . Fourth , each cluster with more than one protein is labeled with a different interacting motif identifier ( iMotif ) , and that iMotif is assigned to all proteins within that cluster . iMotif–iMotif interactions are then derived from interactions in the cluster interaction network where both sides of the interaction have been labeled with an iMotif identifier . For example ( Figure 1 ) , a proteome of six proteins ( namely A , B , C , D , E , and F ) forms a network of interactions that connects proteins A with B , C , and D , and protein E with B , C , D , and F ( step 1 ) . Our method starts by creating a cluster interaction network from the network of protein interactions ( i . e . , six clusters with seven interactions ) ( step 2 ) . Next , the clusters that share the largest number of common interactions are fused ( i . e . , clusters 1 and 5 , with three common interactions , are fused into a new cluster 7 ) . This step is then repeated until the maximum similarity score between the clusters drops below a predefined threshold ( i . e . , N = 2 common interactions ) . Thus , the iterative process will run for another iteration creating a new cluster ( cluster 8 ) by fusing clusters 2 , 3 , and 4 , which have two common interactions ( step 3 ) . Once the iterative process is finished , the method assigns iMotif identifiers to all proteins in clusters with more than one protein ( i . e . , proteins A and E in cluster 7 share iMotif 1 , and proteins B , C , and D in cluster 8 share iMotif 2 ) ( step 4 ) . Moreover , iMotif–iMotif interactions are then derived from the cluster interaction network ( i . e . , one interaction between iMotif 1 and iMotif 2 ) . Figure 2 illustrates iMotif assignments from a network perspective ( Figure 2A ) and from a structural perspective ( Figure 2B ) . A more detailed description of the algorithm is provided as pseudocode in Figure S8 . We have evaluated the method on a test set created by selecting proteins ( i ) with at least five experimentally detected interactions , ( ii ) with at least 80% of their sequence covered by the domains defined in SCOP , and ( iii ) that did not introduce a redundancy bias in the evaluation ( i . e . , if any two sequences had a sequence identity greater than 30% , a BLAST e-value smaller than 10−5 , and the alignment had at least 30 residues , the shortest member of the pair was not selected ) . The final set contained 368 sequences ( Table S6 ) . Due to the restrictions imposed , the test set contains many proteins related to the proteosome and the ribosome . The SCOP family assignment was evaluated by considering as positive assignments those proteins found by the method to have a common iMotif with the query protein . Among these positives , we define as true positives those proteins that have a common SCOP family code with the query protein . Moreover , we define as false negatives the proteins that have the same SCOP family code as the query protein but were not found by the method to share an iMotif . iMotif–iMotif interaction predictions were evaluated against interacting SCOP families obtained from the PDB . Two SCOP domains were considered to interact if they were co-crystallized and had at least two atoms within 5 Å distance . Because we are interested in domain interactions at the protein–protein interaction level , we excluded intrachain interactions from this set . Our method creates a list of putative domain–domain interactions for each predicted iMotif–iMotif interaction by assuming that all domains of the query protein with one iMotif interact with all domains of proteins with the other iMotif . In this context , we define as positive any iMotif–iMotif interaction where the query protein is involved . A positive is then considered a true prediction if at least one of its putative domain–domain interactions is observed in the PDB . Finally , false negatives are interactions observed in the PDB for SCOP families of the query protein that do not appear in any list of putative SCOP family interactions . To avoid biases in the evaluation , only proteins from the test set ( before removing redundancy ) and their SCOP families were considered when counting positives and negatives . The positive predictive value is defined as the number of true positives over the total number of positives , and sensitivity is the number of true positives over the sum of true positives and false negatives . The positive predictive value and sensitivity were calculated with respect to the similarity score threshold used for stopping the clustering . We also define the applicability of the method as the percentage of proteins with at least one positive under a given threshold . For each group of protein sequences with a given iMotif , sequence signatures were generated using the PRATT program [50] , a software tool capable of finding flexible sequence patterns from a set of unaligned sequences . Parameters were set to produce patterns covering a maximum of 15 residues with no more than three consecutive unspecified positions ( gaps ) and a maximum of one flexible ( of variable length ) gap region . The number of nonredundant patterns for all iMotifs was 80 , 654 . Next , all patterns were searched against all proteins in PIANA with at least one interaction ( a dataset of 42 , 040 sequences ) using the ps_scan program [59] . The significance ( i . e . , p-value ) of the association between a sequence pattern and an iMotif was assessed using the binomial distribution , based on the occurrence of the pattern inside the iMotif with respect the whole dataset [60]: where M is the number of protein sequences within an iMotif , n represents the number of proteins within the iMotif that contain the sequence pattern , and p is the probability of finding a protein from the whole dataset that contains the same pattern ( i . e . , the number of proteins containing the pattern divided by the number of proteins in the whole dataset ) . In this work , the best sequence pattern for each iMotif ( Table S5 ) was considered to be the pattern that maximized the number of proteins in the iMotif that had the pattern . Two additional considerations were taken into account for selecting the best pattern for an iMotif: ( i ) the sequence pattern should be found in at least 70% of proteins within the iMotif and ( ii ) the p-value of the pattern should be lower than 10−8 . Interacting motifs were merged by means of an agglomerative hierarchical clustering . Two iMotifs were considered to be similar if they had a common sequence pattern when applying a p-value cutoff of 1 × 10−5 and requiring the pattern to be found in at least 70% of proteins in both iMotifs . Using more stringent p-value cutoffs did not produce any iMotif fusions for proteins from the test set . Hidden Markov Models from the Pfam-A database [61] were assigned to all proteins with at least one known interaction ( a dataset of 42 , 040 sequences ) using the HMMER package [62] . The p-value for each HMM in relation with each iMotif was calculated using the binomial distribution , based on the occurrence of the Pfam inside the iMotif and in the whole dataset ( above ) . All correlations were measured using the Spearman rank correlation coefficient ( rs ) . The assessment of whether two binomial samples of essentiality observations are significantly different was calculated using Fisher's test . The assessment of whether two non-Gaussian samples of evolutionary rate observations come from the same distribution was calculated using the Mann-Whitney U two-sided test . Correlations and differences in the observations were considered significant for p-values lower than 0 . 05 . All tests were performed using the implementation provided by R [63] . | Recent advances in experimental methods have produced a deluge of protein–protein interactions data . However , these methods do not supply information on which specific protein regions are physically in contact during the interactions . Identifying these regions ( interfaces ) is fundamental for scientific disciplines that require detailed characterizations of protein interactions . In this work , we present a computational method that identifies groups of proteins with similar interfaces . This is achieved by relying on the observation that proteins with common interaction partners tend to interact through similar interfaces . The proposed method retrieves protein interactions from public data repositories and groups proteins that share a sensible number of interacting partners . Proteins within the same group are then labeled with the same “interacting motif” identifier ( iMotif ) . The evaluation performed using known protein domains and structural binding sites suggests that the method is better suited for proteins with multiple interacting partners ( hubs ) . Using yeast data , we show that the cellular essentiality of a gene better correlates with the number of interacting motifs than with the absolute number of interactions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"saccharomyces",
"computational",
"biology"
] | 2007 | Characterization of Protein Hubs by Inferring Interacting Motifs from Protein Interactions |
We exploit flow propagation on the directed neuronal network of the nematode C . elegans to reveal dynamically relevant features of its connectome . We find flow-based groupings of neurons at different levels of granularity , which we relate to functional and anatomical constituents of its nervous system . A systematic in silico evaluation of the full set of single and double neuron ablations is used to identify deletions that induce the most severe disruptions of the multi-resolution flow structure . Such ablations are linked to functionally relevant neurons , and suggest potential candidates for further in vivo investigation . In addition , we use the directional patterns of incoming and outgoing network flows at all scales to identify flow profiles for the neurons in the connectome , without pre-imposing a priori categories . The four flow roles identified are linked to signal propagation motivated by biological input-response scenarios .
The nematode Caenorhabditis elegans has been used as a model organism in the life sciences for half a century [1] , and considerable effort has been devoted to elucidate the properties of its nervous system in relation to functional behaviour . The C . elegans connectome was originally charted in 1986 by White et al [2] and has been further refined by analysis and experiments [3] , most recently in the work of Varshney et al [4] . Using experimental techniques such as laser ablations , calcium imaging , optogenetics and sonogenetics , researchers have examined functional properties of individual neurons in connection with motion , learning , or information processing and integration [5–9] . Other studies have quantified the characteristics of the motion of C . elegans , and how these change upon genetic mutations [10–12] . With the increased availability of data from such experiments , there is a need to integrate current knowledge about individual neurons into a comprehensive picture of how the network of neurons operates [2 , 4 , 13] . A number of studies have reported network characteristics of the C . elegans connectome: it is a small-world network [14] satisfying mathematical criteria of efficiency [15] , with a heavy-tailed degree distribution [16] and a core-set of highly-connected , ‘rich club’ neurons [17] . Furthermore , the analysis of modules in the network has shown that certain strongly coupled clusters of neurons can be linked to biological functions [18–22] . Such observations suggest that a system-wide analysis of the connectome can provide valuable functional information . However , finding simplified mesoscale descriptions that coherently aggregate how information propagates in the directed connectome across multiple scales remains a challenge [23] . In this work , rather than focusing on structural features of the network , we analyse the directed and weighted C . elegans connectome from a dynamics-based ( more specifically , flow-based ) perspective . Using dynamics to probe the relationship between the structure and function of a system has become a valuable tool in many settings [23–25] . In particular , dynamics-based approaches have been successfully used to study brain networks ( e . g . , fMRI and DSI data [26–28] ) . For an in depth discussion of network-theoretic methods in neuroscience see the extensive reviews [23 , 29 , 30] . For an overview on dynamical methods for network analysis we refer the reader to Refs . [24 , 25 , 31] and the literature cited therein . Our methods use diffusive processes on graphs as a simple means to link features of the directed network and propagation dynamics . While diffusive flow is a simplification of the actual propagation in the nervous system of C . elegans , we can still gain insight into network properties of dynamical interest [4] . We exploit these ideas in two ways . Firstly , we investigate flow-based partitions of the connectome across multiple scales using the Markov Stability ( MS ) framework for community detection [24 , 31 , 32] . Our analysis detects subgroups of neurons that retain diffusive flows over particular time scales [33] taking into account edge directionality [24 , 34] . We then mimic neuronal ablations computationally , and check all possible single and double ablations in the connectome to detect those that are most disruptive of the flow organisation . Secondly , we extract alternative information of the directed network flows through the Role Based Similarity ( RBS ) framework [35–37] . Without pre-imposing categories a priori , RBS classifies neurons into flow roles , i . e . , classes of neurons with similar asymmetric patterns of incoming and outgoing network flows at all scales , which are directly extracted from the network . Finally , we mimic ‘stimulus-response’ experiments [5 , 7 , 38] , in which signals propagate through the network starting from well-defined sets of input neurons linked to particular biological stimuli . The ensuing time courses of neuronal flows reveal features of information processing in C . elegans , in relation to the obtained flow roles . Our computational analyses are consistent with experimental findings , suggesting that our framework can provide guidance towards the identification of potential neuronal targets for further in vivo experiments .
To reveal the multi-scale flow organisation of the C . elegans connectome , we use the Markov Stability ( MS ) framework described in Sec . ‘A dynamical perspective for community detection in graphs: Markov Stability’ . Conceptually , MS can be understood as follows . Imagine that a drop of ink ( signal ) is placed on a node and begins to diffuse along the edges of the graph . If the graph lacks structural organisation ( e . g . , random ) , the ink diffuses isotropically and rapidly reaches its stationary distribution . However , the graph might contain subgraphs in which the flow is trapped for longer than expected , before diffusing out towards stationarity . These groups of nodes constitute dynamical , flow-retaining communities in the graph , usually signifying a strong dynamic coherence within the group and a weaker coherence with the rest of the network . If we allow the ink to diffuse just for a short time , then only small communities are detected , for the diffusion cannot explore the whole extent of the network . If we observe the process for a longer time , the ink reaches larger parts of the network , and the flow communities thus get larger . By employing dynamics , and in particular by scanning across time , MS can thus detect cohesive node groupings at different levels of granularity [24 , 33 , 39] . In this sense , the time of the diffusion process , denoted Markov time in the following , acts as a resolution parameter . The flow-based community structure of the C . elegans connectome at medium to coarse levels of resolution is shown in Fig 1 . The full scan across all Markov times is shown in S1 Fig and the S1 Dataset . As described above , the partitions become coarser as the Markov time t increases , from the finest possible partition , in which each node forms its own community , to the dominant bi-partition at long Markov times . The sequence of partitions exhibits an almost hierarchical structure , with a strong spatial localisation linked to functional and organisational circuits ( see Fig 2 and S2 Fig ) . These findings are in agreement with the spatial localisation of functional communities often found in brain networks [23] , as well as the hierarchical modularity exhibited by the C . elegans connectome as reported in Ref . [40] . We remark that our community detection method does not enforce a hierarchical agglomeration of communities: the observed quasi-hierarchy and spatial localisation is an intrinsic feature of the C . elegans connectome . In S2 Fig we quantify the deviation of the community structure from a strict hierarchy . At long Markov times , we find robust partitions containing 6 to 2 communities , denoted A to E in Fig 1 . Partition A comprises six communities of varying sizes ( from 9 to 104 neurons ) , well localised along the body of the worm , as seen in Fig 2 ( c . f . Section 2 . 2 in www . wormatlas . org/neuronalwiring . html ) . The two large communities ( A 1 and A 2 ) have head ganglia neurons of all three functional types ( S , I , M ) . In particular , A 1 contains ring motor neurons and interneurons as well as the posterior neurons ALN and PLN , whereas A 2 specifically gathers amphid neurons ( e . g . , AWAL/R , ASKL/R , ASIL/R , AIYL/R ) which feature prominently in the navigation circuit responsible for exploratory behaviour [41] . Communities A 3 , A 4 and A 5 in Partition A consist predominantly of ventral cord motor neurons , differentiated by their soma position along the body ( Fig 2 ) : A 3 contains frontal motor neurons ( e . g . VD1 to VD3 ) ; A 4 consists of mid-body motor neurons ( e . g . VD4 to VD8 ) ; A 5 comprises posterior motor neurons ( e . g . VD9 and VD10 ) . Such partitioning is consistent with the motor neuron segmentation model proposed for C . elegans in Ref . [42] . Finally , A 6 contains highly central neurons such as AVAL/R or PVCL/R , which have been found to belong to a rich-club [17] , as well as interneurons linked to mechanosensation and tap withdrawal functional circuits [20] . The coarser Partitions B and C are quasi-hierarchical merges of A ( Figs 1 and 2 ) . For instance , Partition C has three groupings: head ganglia ( merged A 1 and A 2 ) , frontal motor neurons ( merged A 3 and A 4 ) , and a tail subgroup ( merged A 5 and A 6 ) . Interestingly , at later Markov times , we obtain the distinct , coarser 3-community Partition D , which exemplifies how our method does not enforce a strict hierarchy in the multiscale structure . The three groups in Partition D include a notable community of only three nodes ( interneurons AVFL/R and AVHR ) , which appear as a cohesive group only at this particular timescale . Prominent functional roles of AVF and AVH neurons have been noted previously [4 , 43]: both AVF neurons are responsible for coordination of egg-laying and locomotion [44] . In addition , spectral analyses of the gap-junction Laplacian have shown that AVF , AVH , PHB and C-type motor neurons are strongly coupled [4] . Finally , the two communities in the coarsest Partition E split the connectome anatomically into a group with head and tail ganglia ( red ) , and another group predominantly with motor neurons ( cyan ) . Laser ablation experiments are invaluable to probe the functional role of neurons [5–7] , but are time consuming and technically challenging . We have used our computational framework to assess the effect that an ablation of a single neuron , or of a pair of neurons , has on the signal flow in the connectome . To this end , we compare the flow-based partitions obtained for the ablated connectome against the original network . If an ablation creates large distortions in the flow structure , the partitions of the ablated network will change drastically or become less robust compared to those found in the unablated network . We have carried out a systematic computational analysis of all single and double neuron ablations in the connectome . A complementary analysis of the directed connectome of C . elegans is provided by the Role Based Similarity ( RBS ) framework [35 , 36] , which identifies groups of nodes with similar flow profiles in the network without imposing a priori the type or number of groups . Such groups of neurons display the same character ( or flow role ) in terms of their role in the generation , distribution and consumption of flow in the network . Briefly , RBS obtains a flow profile for each node from its incoming and outgoing flows at all scales . We then group the nodes into classes ( ‘flow roles’ ) with similar in- and out-flow patterns . Because they include information at all scales , flow roles capture nuanced information about the network , beyond pre-defined categories ( e . g . , sources , sinks , hubs ) or combinatorial notions based on immediate neighbourhoods ( e . g . , roles from Structural Equivalence [49] and Regular Equivalence [50] ) . Details of the RBS methodology are given in Refs . [34–37] , and summarised in Section ‘Finding flow roles in networks: Role-based similarity’ and in the S4 Fig . In the C . elegans connectome , we identify four distinct classes of neurons according to their flow profiles ( Fig 6 ) . These flow roles are distinct from the groupings into communities ( see an analysis of communities and their mix of flow roles in the S5 Fig ) . Two of the roles ( R1 and R2 ) have a dominant ‘source’ character ( i . e . , higher average in-degree than out-degree ) and contain most of the nodes with high PageRank ( S6 Fig ) . The other two roles ( R3 and R4 ) have a dominant ‘sink’ character and nodes with low PageRank . Note , however , that these roles are not just defined by average properties , but by their global flow patterns in the network . As seen in Fig 6b , R1 is upstream from R3 and R4 , whereas R2 is mostly upstream from R4 . Furthermore , R4 is an almost pure downstream module , whereas R3 has a stronger feedback connection with R1 . The RBS flow roles are linked to physiological properties of the neurons ( Fig 6c and 6d ) . R4 corresponds to a group of motor neurons ( mostly ventral chord motor neurons ) consistent with its downstream character , whereas R1 is a group of mostly sensory and inter-neurons with heavy localisation in the head . R3 is a group with a balanced representation of all three types of neurons ( including some polymodal neurons ) localised in the head . Indeed , most ring neurons in R3 are in community A 1 , indicating a self-contained unit that process head-specific behaviour , such as foraging movements and the head withdrawal reflex [45] . Our RBS analysis also reveals a specific flow profile ( R2 ) containing 13 neurons ( mainly sensory and interneurons , mostly upstream from the motor neurons in R4 ) , the majority of which are responsible for escape reflexes triggered in the presence of noxious factors ( Table 2 ) . This group can be seen as a group of escape response neurons and include: the PVDL/R neurons , which sense cold temperatures and harsh touch along the body; FLPL/R , which perform the equivalent task for the anterior region; PHB neurons responsible for chemorepulsion; PHCR , which detects noxiously high temperatures in the tail; SDQL and PQR , which mediate high oxygen and CO2 avoidance , respectively; and PLMR , a touch mechanosensor in the tail [2] . This escape response group is heavily over-connected to command neurons AVAL/R , AVDL/R , DVA , PVCL/R , all of which modulate the locomotion of the worm . ( Specifically , there are 48 connections from R2 to these particular command neurons in contrast to the ∼12 connections expected at random . ) Note that AVDL/R and DVA are in R1 , whereas AVAL/R and PVCL/R are in R4; the R2 group thus links directly to motor locomotion neurons across the worm . We remark that this group of neurons was found exclusively through the analysis of their all-scale in/out flow profiles , without any other extrinsic information . Despite its modest size , the nervous system of C . elegans can sense and react to a wide range of mechanical , chemical and thermal factors [45] . Standard notions in neuroscience hold that stimuli lead to motor action due to information progressing from sensory through inter- to motor neurons [51] . However , the underlying mechanisms and precise signal flows are still far from understood . In the absence of measurements probing such pathways , and as a first approximation to more realistic nonlinear dynamical models , we use here simplified diffusive dynamics ( see Section ‘Propagation dynamics in the network’ ) to mimic signal propagation in the C . elegans directed network . Such an approach , already suggested by Varshney et al . [4] , is naturally linked to MS multiscale community detection and to the identification of RBS flow roles , since both Markov Stability and Role Based Similarity are intrinsically defined in terms of a diffusive process on the graph . To mimic the propagation of stimuli associated with particular biological scenarios , a normalised initial flow vector ϕ ( 0 ) is localised at specific input neurons and we observe the decay towards stationarity under Eq ( 5 ) : θ ( t ) = ϕ ( t ) - π . ( 1 ) We also define q ( t ) , which will be used to detect overshooting neurons: q i ( t ) = ϕ i ( t ) π i = 1 + θ i ( t ) π i . ( 2 ) Initially , θi ( 0 ) is positive only for the input neurons where we inject the signal , and negative for all other neurons . Asymptotically , the vector of flows ϕ ( t ) approaches the stationary solution π , and θi ( t ) → 0 , ∀i . However the approach to the stationary value can be qualitatively different . In some cases , θi ( t ) can become positive , if neuron i receives an influx of flow that drives it to ‘overshoot’ above its stationary value; in other cases , neurons approach stationarity without overshooting . The different behaviour depends on the particular initial input and the relative location of each neuron in the network . Motivated by several experimental studies , we have conducted four case studies corresponding to different biological scenarios in which the input is localised on specific neurons: We exemplify the procedure in detail through the posterior mechanosensory stimulus ( i1 ) , but detailed results for the other stimuli are provided in the S8 , S9 and S10 Figs . As shown in Fig 7a , the signal proceeds ‘downstream’ following the expected biological information processing sequence , S→I→M . The signal is initially concentrated on the input neurons ( mostly sensory ) ; then propagates out primarily to interneurons , which overshoot and peak at t ≈ 1 . 5; and is then passed on to motor neurons , which slowly increase towards their stationary value . The flow roles obtained above provide further insight into the propagation of stimuli . As seen in Fig 7c , the input for the tail mechanosensory scenario ( i1 ) is heavily concentrated on R2 neurons ( the escape response group ) , from which the signal flows quickly towards the other upstream ( head ) group R1 , followed by propagation towards the downstream group R4 . Finally , the signal spreads more slowly to R3 , the head-centric downstream unit . This pattern of propagation carries onto the sequence of strong response neurons ( Fig 7b ) , and reflects the fact that R2 contains posterior upstream units , and mirrors the strong connectivity of R2 with motor neurons in R1 ( AVDL/R and DVA ) and R4 ( PVCL/R ) , as discussed above . To detect key neurons comprising the specific propagation pathways , we find strong response neurons , i . e . , those with large overshoots relative to their stationary value , q max , i = max t q i ( t ) > 1 + 2 3 . See S7 Fig for a full description of the procedure . According to this criterion , we obtain 26 strong response neurons for scenario ( i1 ) . The neurons have large overshoots in two time windows after the inital input ( Fig 7b ) . The details of the signal propagation ( Fig 7d ) show that a first wave of peak responses ( around t ≈ 1 ) corresponds mostly to overshooting interneurons , including AVDL/R and DVA , responsible for mechanosensory integration , and PVCL/R , drivers of forward motion [5 , 45] . The second wave of peaks ( around t ≈ 3 ) contains predominantly ventral B-type motor neurons , e . g . , DB2-7 and VB11 . Such B-type motor neurons are responsible for forward motion . Hence the progression of overshooting neurons suggests a plausible biological response for a posterior mechanosensory stimulus [7 , 45] . The overshooting behaviour of the neurons is not captured by other static measures of the network ( e . g . , in/out degree or pagerank ) , as shown in S12 Fig .
We have presented an integrated network-theoretic analysis of the C . elegans connectome in terms of directed flows . We exploit the connection between diffusive processes and graph-theoretical properties , which intimately links structure and dynamics , to elucidate dynamically relevant features in the connectome . Although diffusive processes are a coarse approximation of physiological signal propagation , they can be used to extract systemic dynamical features , specifically in the case of non-spiking neuronal systems such as C . elegans [4] . Using the Markov Stability ( MS ) framework , we have identified flow-based groupings of neurons in the C . elegans connectome at different levels of granularity . Previous studies [20–22] have aimed at uncovering modules based on structural properties of the network , usually considering a particular scale so as to find one partition ( e . g . , modularity at the standard resolution ) . In S1 Text we provide a detailed comparison of MS multiscale flow structures against partitions found by modularity [20 , 21] , stochastic block models [22] and MapEquation [52] . The partitions obtained by MS at a particular scale are closer to those obtained with directed modularity . The MS framework , however , provides a multiscale description across all scales by sweeping the Markov time [33] , respecting and exploiting directionality . In doing so , it reveals an intrinsic , quasi-hierarchical organisation of the connectome , giving insight into relevant features of signal propagation . The partitions found by MS are in good agreement with C . elegans physiology , and summarise previously observed features , such as the hierarchical and spatial organisation of neuronal communities [23 , 40] . The obtained flow-based organisation highlights the prominent position of particular neurons , such as AVF and AVH , and allows for a systematic exploration of single and double ablations most disruptive of signal flows , thus providing insight into candidate neurons for further experimental investigations . Examples of such neurons include , among others: the synergistic effects caused by neuron AIAR in double ablations; the global role of D-type motor neurons , which often appear as relevant in single ablations; or the role of polymodal ( I/M ) SAAVL/R head neurons [45] , about which little is known but which appear in the R2 group and are salient in our ablations . Several other examples are discussed in the text , and further such hypotheses may be formulated based on the full set of ablation scores we provide in S1 Data as a resource to experimentalists investigating the physiology of particular neurons . Other methods can be used to study the effect of ablations using , for example , measures of centrality , efficiency or information transfer [53 , 54] . Our study of ablations gives distinct results , as shown in S3 Fig . For instance , because our measures focus on the disruption of the flow community structure at different scales , our approach can provide a structured view of the effect of ablations for different neuron types , as shown in Figs 3–5 . As a complementary flow-based perspective , we have used Role Based Similarity to identify classes of neurons with similar patterns of flow in the C . elegans nervous system . Rather than reflecting any measure of connectedness in the network , such flow roles ( or flow profiles ) reflect similar roles in the generation , distribution and consumption of flow in the directed connectome . In previous work , neurons have been assigned to roles by exploring the core-periphery structure [55] , or by examining the connections of nodes within and between communities [47 , 56] . Other notions of roles have been based on the use of centrality scores , or on combinatorial notions of social neighbourhoods , as in regular and structural equivalence [49 , 50] . RBS takes a different approach by grouping neurons according to their patterns of in/out flows at multiple scales in the graph , irrespective of their community membership and going beyond standard classifications [34 , 37] . See S2 Text and S6 Fig for a comparison of RBS flow roles , regular equivalence and community roles . The RBS analysis of flow profiles finds two groups of mostly upstream neurons and two groups of mostly downstream neurons , yet with a specific inter-connectivity pattern . In particular , the analysis singles out a small group of upstream neurons ( R2 ) , which is functionally related to escape responses from noxious factors , and could also be the object of further experimental investigation . The RBS roles are also informative in conjunction with signal propagation from ‘input-response’ in silico biological scenarios ( see S11 Fig ) . In particular , the R2 group plays an important role in posterior biological stimuli , channelling stronger and faster responses , whereas R3 ( the downstream , head-centric group ) constitutes a self-contained set of neurons mainly accessible via the upstream , head-centric R1 group . Therefore , the propagation profiles obtained for different biological scenarios suggest a graded organisation of the roles of nodes in terms of upstream-downstream information , which could provide valuable insight into functional circuits . Interesting theoretical extensions of the current work would include considering the C . elegans connectome as a multiplex network; taking into account the different types of synapse in a more explicit fashion; and enriching the dynamics of the model by incorporating the effects of inhibitory synapses and nonlinearities in the dynamics . Furthermore , one may explore more intricate dynamics by incorporating the memory of information flow using higher order Markov models [57 , 58] . Our computational tools could be used in conjunction with experimental techniques , as an aid to the generation of functional hypotheses for experimental evaluation . With the eventual aim of linking wiring properties of the connectome with information processing and functional behaviour , high throughput experiments ( e . g . , systematic ablation of several neurons ) coupled with advancements in neuronal monitoring that can allow recordings from thousands of neurons simultaneously [59] could deliver time course measurements to characterise signal propagation in relation to function . Another interesting area of future work would be the evaluation of ablation and propagation scenarios as related to quantitative behavioural investigations upon more general ablational/mutational strategies in C . elegans [10–12] , as well as comparative studies of the flow architecture in different nematode species [60] . Such comparative analyses between the functional and structural network of the connectome could yield valuable information in bridging the relation between structure and function in network neuroscience .
The information of the large component of the connectome network is encoded into the n × n adjacency matrix A ( n = 279 ) , where entry Aij counts the total number of synapses ( both chemical synapses and gap junctions ) connecting neuron i to neuron j [4] . Note that chemical synapses are not necessarily reciprocal , hence A ≠ AT . Therefore the connectome is a directed , weighted network . The network is relatively sparse , with 2990 edges: 796 edges formed by gap junctions only; 1962 containing only chemical synapses; 232 edges with both gap junctions and chemical synapses present . The vector of out-strengths , which compiles the sum of all synapses for each neuron , is d = A1 ( where 1 is the n × 1 vector of ones ) . The average out-strength per neuron is 29; ranging from the maximum ( 256 ) attained by neuron AVAL to the minimum ( 0 ) attained by the motor neuron DD6 , which is the only sink in the network . The network is not strongly connected . Methods with different levels of complexity have been used to study signal propagation in the C . elegans connectome ( see , e . g . , Refs . [4 , 51 , 61–63] ) . Here , we use a continuous-time diffusion process as a simple proxy for the spread of information in this neuronal network . Note that gap junctions may be simply modelled as linear resistors and , although chemical synapses are likely to introduce nonlinearities , their sigmoidal transfer functions may be well approximated by a linearisation around their operating point . Indeed , as remarked by Varshney et al . [4] , such an approach has additional merit in C . elegans , where neurons do not fire action potentials and have chemical synapses that release neurotransmitters tonically [64] . Thus , linear systems analysis is in this case an appropriate tool that can provide valuable insights [4] . Interestingly , athough simplified , such linear models have been successfully applied even to the analysis of spatio-temporal behaviour of strongly nonlinear neuronal networks [65] . The signal on the nodes at time t is represented by the 1 × n row vector ϕ ( t ) governed by the differential equation d ϕ d t = ϕ M - I , ( 3 ) where I is the identity matrix and M is the transition matrix defined as follows: M = τ D † A + 1 n 1 - τ 1 + τ 1 d i = 0 1 T . ( 4 ) Here , τ ∈ ( 0 , 1 ) is the Google teleportation parameter ( and we take τ = 0 . 85 as is customary in the literature ) ; 1 d i = 0 is the indicator vector of sink nodes; and the diagonal matrix D† is the pseudo-inverse of the degree matrix: D i i † = 0 if d i = 0 1 / d i if d i ≠ 0 . The matrix M describes a signal diffusion along the directed edges with an additional re-injection of external ‘environmental noise’: each node receives inputs from its neighbours ( which transmit flow along their outgoing links according to their relative weight with probability τ ) and receives a constant external re-injection of size ( 1 − τ ) /n . For pure sinks , the outgoing flow is uniformly redistributed to all nodes so as to avoid the signal accumulating at nodes with no out-links . Mathematically , this reinjection of probability ( known as teleportation in the networks literature ) guarantees the existence of a unique stationary solution for Eq ( 3 ) , even when the network is not strongly connected [24 , 66] . Biophysically , the teleportation can be understood as modelling the random interactions with the external environment . Let ϕ ( 0 ) be the input , i . e . , the signal at t = 0 . The solution of Eq ( 3 ) is then: ϕ ( t ) = ϕ ( 0 ) exp t [ M - I ] , ( 5 ) with stationary solution ϕ ( t → ∞ ) = ( ϕ ( 0 ) ⋅ 1 ) π , where π is the dominant left eigenvector of M , known as PageRank [66] . Therefore , under a unit-normalised input , ϕ ( t ) ⋅ 1 = 1 ∀t , and the stationary solution is π . The diffusive dynamics Eq ( 3 ) can be exploited to reveal the multiscale organisation of the C . elegans connectome using the Markov Stability community detection framework [24 , 31 , 32] . Markov Stability finds communities across scales by optimising a cost function related to this diffusion ( parametrically dependent on time ) over the space of all partitions . More formally , a partition P of the n nodes of the network into m non-overlapping communities is encoded as a n × m indicator matrix H P: [ H P ] i c = 1 if node i belongs to community c 0 otherwise . ( 6 ) Given a partition matrix H P , we define the time-dependent clustered autocovariance matrix: R ( t , H P ) = H P T Π exp ( t [ M - I ] ) - π π T H P , ( 7 ) where Π = diag ( π ) . The matrix entry [ R ( t , H P ) ] c f quantifies how likely it is that a random walker starting in community c will end in community f at time t , minus the probability for such an event to happen by chance . To find groups of nodes where flows are trapped more strongly over time t than one would expect at random , we find a partition P that maximises r ( t , H P ) = trace R ( t , H P ) . ( 8 ) We define r ( t , H P ) as the Markov Stability of partition P at time t [24 , 32] . Maximising r ( t , H P ) over the space of all partitions for each time t results in the sequence of optimal partitions: P max ( t ) = arg max P r ( t , H P ) . ( 9 ) Although the optimisation Eq ( 9 ) is NP-hard , there exist efficient heuristic algorithms that work well in practice . In particular , it has been shown that this optimisation can be carried out using any algorithm devised for modularity maximisation [24 , 31 , 32] . In this work , we use the Louvain algorithm [67] , which is known to offer high quality solutions whilst remaining computationally efficient . The code for Markov Stability can be found at github . com/michaelschaub/PartitionStability . As an additional improvement of the optimisation of P max ( t ) , we run the Louvain algorithm ℓ = 100 times with different random initialisations for each Markov time t , and generate an ensemble of solutions { P i ( t ) } i = 1 ℓ . From this ensemble , we pick the best partition P ^ ( t ) according to our measure Eq ( 8 ) : max i { P i ( t ) } i = 1 ℓ ↦ P ^ ( t ) ≈ P max ( t ) . Ideally , the optimised partition from the ensemble , P ^ ( t ) , will be close to the true optimum , P max ( t ) . To identify the important partitions across time , we use the following two robustness criteria [33 , 68]: To mimic in silico the ablation of neuron i , we remove the i-th row and column of the adjacency matrix A , and analyse the change induced in the Markov Stability community structure of the reduced ( n − 1 ) × ( n − 1 ) matrix A[i] . Double ablations are mimicked by simultaneously removing two rows ( and their corresponding columns ) to obtain the reduced ( n − 2 ) × ( n − 2 ) matrix A[i , j] . In directed networks , nodes can have different ‘roles’ , e . g . , sinks , sources or hubs . In complex directed networks , functional roles may not fall into such simple categories , yet nodes can still be characterised by their contribution to the diffusion of in- and out-flows . Here we use a recent method ( Role-Based Similarity , RBS ) to uncover roles in directed networks based on the patterns of incoming and outgoing flows at all scales [35 , 36] . The main idea underpinning RBS is that nodes with a similar in/out flow profile play a similar role , regardless of whether they are near or far apart in the network . Each node is associated with a feature vector xi containing a weighted number of in- and out-paths of increasing lengths beginning and ending at the node . The feature vectors are collected in the feature matrix X: X = x 1 ⋮ x n = ⋯ ( β A T ) k 1 ⋯ ︷ paths in | ⋯ ( β A ) k 1 ⋯ ︷ paths out , ( 17 ) where β = α/λ1 , with λ1 the spectral radius of the adjacency matrix A and α ∈ ( 0 , 1 ) . The cosine between feature vectors gives the similarity score between nodes: Y i j = x i x j T ‖ x i ‖ 2 ‖ x j ‖ 2 . ( 18 ) The n × n matrix Y quantifies how similar the directed flow profiles between every pair of nodes are . Nodes with identical connectivity have Yij = 1 , whereas in the case of nodes with dissimilar flow profiles ( e . g . , if i is a source node with no incoming connections and j is a sink node with no outgoing connections ) , then their feature vectors are orthogonal and Yij = 0 . As outlined in Refs . [35–37] , we compute the similarity matrix Y iteratively with α = 0 . 95 , and apply the RMST algorithm to obtain a similarity graph , in which only the important information of Y is retained . We then extract flow roles in a data-driven manner without imposing the number of roles a priori by clustering the similarity graph ( see S4 Fig ) . The flow roles so obtained have been shown to capture relevant features in complex networks , where other role classifications based on combinatorial concepts and neighbourhoods fail [34 , 37] . In particular , our flow roles are fundamentally different from notions of roles in social networks based on Structural Equivalence [49] and Regular Equivalence [50] . Such equivalence measures do not incorporate information about the large scales of the network and are sensitive to small perturbations , making them unsuitable for complex networks such as the C . elegans connectome [34] ( see S6 Fig for roles based on Regular Equivalence ) . | One of the goals of systems neuroscience is to elucidate the relationship between the structure of neuronal networks and the functional dynamics that they implement . An ideal model organism to study such interactions is the roundworm C . elegans , which not only has a fully mapped connectome , but has also been the object of extensive behavioural , genetic and neurophysiological experiments . Here we present an analysis of the neuronal network of C . elegans from a dynamical flow perspective . Our analysis reveals a multi-scale organisation of the signal flow in the network linked to anatomical and functional features of neurons , as well as identifying different neuronal roles in relation to signal propagation . We use our computational framework to explore biological input-response scenarios as well as exhaustive in silico ablations , which we relate to experimental findings reported in the literature . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"infographics",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"nervous",
"system",
"caenorhabditis",
"electrophysiology",
"neuroscience",
"animals",
"motor",
"neurons",
"animal",
"models",
"caenorhabditis",
"elegans",
"model",
"organisms",
... | 2016 | Flow-Based Network Analysis of the Caenorhabditis elegans Connectome |
Overexpression or mutation of α-Synuclein is associated with protein aggregation and interferes with a number of cellular processes , including mitochondrial integrity and function . We used a whole-genome screen in the fruit fly Drosophila melanogaster to search for novel genetic modifiers of human [A53T]α-Synuclein–induced neurotoxicity . Decreased expression of the mitochondrial chaperone protein tumor necrosis factor receptor associated protein-1 ( TRAP1 ) was found to enhance age-dependent loss of fly head dopamine ( DA ) and DA neuron number resulting from [A53T]α-Synuclein expression . In addition , decreased TRAP1 expression in [A53T]α-Synuclein–expressing flies resulted in enhanced loss of climbing ability and sensitivity to oxidative stress . Overexpression of human TRAP1 was able to rescue these phenotypes . Similarly , human TRAP1 overexpression in rat primary cortical neurons rescued [A53T]α-Synuclein–induced sensitivity to rotenone treatment . In human ( non ) neuronal cell lines , small interfering RNA directed against TRAP1 enhanced [A53T]α-Synuclein–induced sensitivity to oxidative stress treatment . [A53T]α-Synuclein directly interfered with mitochondrial function , as its expression reduced Complex I activity in HEK293 cells . These effects were blocked by TRAP1 overexpression . Moreover , TRAP1 was able to prevent alteration in mitochondrial morphology caused by [A53T]α-Synuclein overexpression in human SH-SY5Y cells . These results indicate that [A53T]α-Synuclein toxicity is intimately connected to mitochondrial dysfunction and that toxicity reduction in fly and rat primary neurons and human cell lines can be achieved using overexpression of the mitochondrial chaperone TRAP1 . Interestingly , TRAP1 has previously been shown to be phosphorylated by the serine/threonine kinase PINK1 , thus providing a potential link of PINK1 via TRAP1 to α-Synuclein .
Parkinson's disease ( PD ) is the second most prevalent neurodegenerative disease behind Alzheimer's disease ( AD ) , with an incidence rate of approximately 110–300 per 100 , 000 persons above the age of 50 [1] . The movement disorder is characterized by the selective death of dopaminergic neurons in the substantia nigra pars compacta ( SNc ) [2] . Death of SNc neurons results in a reduction of dopamine ( DA ) levels within their key efferent target , the striatum [3] . Mitochondrial Complex I activity deficit and evidence of enhanced oxidative stress within affected brain regions are also observed in PD [4]–[6] . Age and pesticide/herbicide exposure are the most important disease risk factors [7]–[9] . Importantly , there is no clinical therapy available that has been shown to slow or reverse PD . While the majority of PD is diagnosed as idiopathic , 5–10% of cases are attributable to familial forms of PD [10] . Although genetic PD represents only a small percentage of patients , mutations in these genes point to underlying biochemical pathways that could also be relevant to sporadic PD patients . Three missense mutations in the small pre-synaptic protein α-Synuclein ( SNCA/PARK1/4; GenBank ID 6622 ) have been shown to result in autosomal-dominant PD . A critical effect of protein dose on pathology is implicated by disease-causing gene duplication and triplication [11]–[14] . α-Synuclein is also a major protein component of the Lewy Bodies ( LB ) , the key histologic feature of dopaminergic and non-dopaminergic neurons found in PD patients [15] . Thus , α-Synuclein is strongly suggested to be a causal factor in PD pathogenesis . Human α-Synuclein mutation or overexpression results in cytotoxicity , with [A53T]α-Synuclein being the most toxic variant known . Direct cell loss can be induced in both in vitro and in vivo models of yeast , C . elegans , Drosophila , rat , mouse , and non-human primate [16]–[23] . The formation of α-Synuclein oligomers from their native unfolded state is linked to cell membrane damage and results in dysfunction of multiple cell systems such as the ubiquitin proteasome system , the endoplasmic reticulum and lysosomes [24]–[31] . Recent data also suggests that α-Synuclein plays a role in modulating both mitochondrial function and damage . α-Synuclein-overexpressing cells exhibit multiple markers of mitochondrial dysfunction , including increased protein oxidation , increased ROS production , loss of mitochondrial membrane potential and reduced Complex I activity [32]–[38] . Several groups have demonstrated that α-Synuclein's entry in mitochondria is mediated via an N-terminal mitochondrial targeting sequence , with localization at the inner membrane [37] , [38] . Moreover , PD patient brain histology shows α-Synuclein accumulation within mitochondria of the SNc and striatum , a feature absent in control brains [38] . Mitochondrial dysfunction associated with adenosine triphosphate ( ATP ) depletion and electron transport chain ( ETC ) defects reduces the cell's ability to handle oxidative protein damage and cellular tasks , suggesting a possible reason for cell death . In PD patient brains , early DA reduction indicates the withdrawal of SNc striatal projections , finally resulting in DA neuron loss and PD-related symptoms of rigidity and akinesia . In flies , expression of [A53T]α-Synuclein is accompanied by an age-dependent loss of DA and DA neurons , respectively . Thus , fly head DA levels provide an indirect readout for [A53T]α-Synuclein-induced toxicity . To further investigate the mechanism and identify novel modifiers of α-Synuclein toxicity , we performed a genome-wide genetic screen in Drosophila . In this screen , we identified , among other gene products , the mitochondrial chaperone protein TRAP1 ( GenBank ID 10131 ) as a novel modifier of [A53T]α-Synuclein-induced DA loss . TRAP1 has previously been shown to function downstream of the PD-related serine/threonine kinase PINK1 ( GenBank ID 65018 ) . PINK1-induced phosphorylation of TRAP1 seems to be necessary for the protein's protective effects against oxidative stress [39] . In our report we further characterize the functional consequences of TRAP1 reduction or overexpression in Drosophila , in primary neurons and dopaminergic cell lines and the effects on mitochondrial morphology and function .
Expression of α-Synuclein in Drosophila is established as a useful model of PD [21] . As the fly lacks an α-Synuclein homolog , this model relies on ectopic expression of human α-Synuclein using the UAS/GAL4 system [40] . We have previously analyzed DA neuron number in aged flies , expressing different mutant variants of α-Synuclein . Compared to controls , wild type α-Synuclein did not cause a decline in DA neuron number . Moreover , locomotion was not impaired in aged wild type α-Synuclein-expressing flies [41] . Based on these results , we chose [A53T]α-Synuclein for our screening . With single copy expression of a UAS:[A53T]α-Synuclein transgene ( A53T ) in aminergic neurons ( dopa decarboxylase-GAL4 driver , ddc-GAL4 ) ( Figure 1A ) , no difference to overall fly fitness , as assessed by longevity , was observed ( Figure 1B ) . In contrast , expression of two transgene copies , resulting in higher expression levels ( Figure 1A ) , caused earlier lethality compared to controls ( Figure 1B ) . However , when we measured DA levels of flies expressing one copy of [A53T]α-Synuclein under control of ddc-GAL4 ( ddc>A53T ) , we noticed a significant decrease in DA levels with aging ( Figure 1C ) . Thus , measuring DA levels using high performance liquid chromatography represents a sensitive system to address DA levels in fly heads . After carefully addressing sensitivity , specificity and reproducibility of our readout marker ( Figure 1D , Figure S1 ) , we decided to perform a genome-wide screen to identify modifiers of [A53T]α-Synuclein-induced DA loss in vivo . Thus , flies with expression of [A53T]α-Synuclein in aminergic neurons were crossbred to fly lines carrying chromosomal deletions ( deficiencies ) , utilizing the “Bloomington Deficiency Kit” . Progeny were screened for changes in DA loss over time ( a summary of the screen results is given in Text S1 and Tables S1 , S2 , S3 , S4 ) . Although detailed single gene analysis is still ongoing , we identified a large number of genes coding for proteins involved in mitochondrial function within the candidate deficiencies . Therefore , we additionally cross-referenced our data with results from a genome-wide RNAi-screen , set to identify modulators of mitochondrial function [42] . Common genes were screened for alteration of [A53T]α-Synuclein-toxicity with respect to viability and DA loss . Among the deficiencies screened , Df ( 2R ) nap9 caused the greatest enhancement of [A53T]α-Synuclein-induced DA loss of all non-lethal interacting deficiencies . Of the 153 genes deleted by Df ( 2R ) nap9 , we found TRAP1 reduction to enhance [A53T]α-Synuclein-induced DA loss . The loss-of-function allele TRAP1[KG06242] ( hereafter referred to as TRAP1[KG] , Figure S2 ) caused a reduction of fly head DA levels similar to those of Df ( 2R ) nap9 ( data not shown ) . However , TRAP1[KG] did not alter DA levels ( Figure 1D ) . Thus , flies with reduced TRAP1 in combination with the ddc-GAL4 driver ( TRAP1[KG]/+; ddc/+ ) served as controls in later analysis . TRAP1 is a mitochondrial chaperone , recently reported as a downstream phosphorylation target of the PD protein PINK1 in rat and human cell lines [39] . As both fly and human TRAP1 share high sequence homology , we generated a UAS-transgenic fly to express human TRAP1 ( hTRAP1 ) . Interestingly , overexpression of hTRAP1 in fly heads was able to provide a rescue effect against [A53T]α-Synuclein-induced DA loss ( Figure 2A ) . The effects of TRAP1 on DA levels were also reflected by its effect on tyrosine hydroxylase ( TH ) -positive neurons . ddc>A53T expressing flies showed increased loss of TH-positive neurons if TRAP1 levels were reduced ( Figure 2B ) . In contrast , hTRAP1 overexpression was able to restore [A53T]α-Synuclein-induced loss of TH-positive neurons to control levels . Interestingly , ddc-driven overexpression of hTRAP1 did not increase longevity of ddc>A53T heterozygous flies ( not shown ) . In PD patients , a reduction in brain DA content is later followed by neuronal decline . The same seems to hold true for flies . Although no reduction in longevity of ddc>A53T/+ flies is observed , these flies display a significant reduction in DA content . A more pronounced DA reduction ( ddc>A53T/ddc>A53T ) results in neuronal decline , eventually leading to early death , reflected by a significantly shortened lifespan ( Figure 1B ) . This might explain why TRAP1 will still provide 100% protection against loss of neurons ( Figure 2B ) , even if DA levels have already started to decline ( Figure 2A ) . PD is clinically defined as a movement disorder . Thus , key to an animal disease model recapitulating this phenotype is loss of locomotor ability . Locomotion in flies is measurable using the negative geotaxis assay . In agreement with previous reports , ddc>A53T flies showed an age-related deficit in climbing ability [21] , [43] . Notably , ddc-driven hTRAP1 expression was able to significantly rescue the locomotion deficit in ddc>A53T flies ( Figure 2C ) . Therefore , taken together , these data indicate that the rescue of head DA content is sufficient to restore motor ability . Our data suggest that TRAP1 protects from toxic effects induced by [A53T]α-Synuclein . However , TRAP1 might also provide protection to any toxic insult . To address this possibility , we examined if TRAP1 provides protection against toxicity induced by the expression of other well-known toxic proteins/peptides . Eye-specific expression ( GMR-GAL4 ) of either human Tau or SCA3-derived polyglutamine stretches ( PolyQ ) in the fly eye result in a rough eye phenotype ( REP ) . These REPs are sensitive to genetic modifiers and have successfully been used for screening [44] , [45] . Interestingly , overexpression of TRAP1 did not show a rescue of the PolyQ-induced REP . Moreover , silencing of TRAP1 by RNAi did not enhance the REP ( Figure 2D ) . Similar results were obtained with Tau-expressing flies ( not shown ) . A general protective role of TRAP1 in any toxic trigger is therefore unlikely but appears to be specific for α-Synuclein toxicity . In flies , overexpression of hTRAP1 was able to reduce [A53T]α-Synuclein-induced loss of DA in fly heads and loss of DA neurons , respectively . Thus , to additionally confirm that overexpression of hTRAP1 is able to rescue [A53T]α-Synuclein-induced sensitivity in vertebrate neurons , we used terminally-differentiated rat primary neuron cultures . Lentiviral infection specificity and efficacy in these cells was first verified ( Figure 3A , 3B ) . As the neurons did not display robust toxicity upon [A53T]α-Synuclein expression alone , cells were exposed to low doses of the mitochondrial Complex I inhibitor rotenone . Compared to GFP-virus infected cells ( control ) , co-expression of [A53T]α-Synuclein significantly enhanced cell loss ( Figure 3C ) . In agreement with fly data , coincident overexpression of TRAP1 restored survival to control values . Interestingly , expression of TRAP1 alone enhanced survival beyond that of control cells , indicative of a protective effect of TRAP1 on neurons independent of effects on [A53T]α-Synuclein-induced toxicity . To study the functional role of TRAP1 , we used both TRAP1 overexpression and specific knockdown by small interfering RNA ( siRNA ) in human embryonic kidney cells-293 ( HEK293 ) cells . Two different siRNAs directed against TRAP1 were first compared for efficacy . Both were able to reduce endogenous TRAP1 expression in HEK293 cells ( Figure S3 ) . The most efficient siRNA was used for all further investigations . To confirm whether treatment of HEK293 cells mimicked the in vivo fly and in vitro rat neuron data concerning TRAP1 and stress sensitivity , HEK293 cells were treated overnight with a low dose of either hydrogen peroxide ( Figure 4A ) or the Complex I inhibitor rotenone ( Figure 4B ) . [A53T]α-Synuclein expression enhanced cell sensitivity to both stressors . Reduction in TRAP1 expression further reduced survival in the presence of [A53T]α-Synuclein . For both , rotenone and hydrogen peroxide treatment , overexpression of TRAP1 in the context of [A53T]α-Synuclein expression attenuated the decrease in cell survival . The magnitude of the rescue effect was greatest when cells were exposed to rotenone ( Figure 4A , 4B ) . In cells without [A53T]α-Synuclein expression , reduction of TRAP1 also caused stress sensitization . These data corroborate the reported function of TRAP1 as a protective mitochondrial chaperone [46]–[48] . Previous reports have indicated that [A53T]α-Synuclein may interfere with mitochondrial respiration , in particular with Complex I function [34] . Given the noted rescue effect of TRAP1 on rotenone-treated cells with or without [A53T]α-Synuclein expression , we hypothesized that the TRAP1 effect on [A53T]α-Synuclein may in part be related to altered ETC function . Thus , ATP production via Complex I was assayed in cells without oxidative stress , to evaluate the general effects of [A53T]α-Synuclein on ETC in combination with altered TRAP1 levels . Expression of [A53T]α-Synuclein reduced Complex I activity in HEK293 cells ( Figure 4C ) . TRAP1-silencing enhanced this reduction , while TRAP1 overexpression rescued the [A53T]α-Synuclein-induced defect . In light of the defects observed in [A53T]α-Synuclein-induced Complex I ATP production ( Figure 4C ) , total ATP levels in the cell were also investigated . Only cells expressing [A53T]α-Synuclein in combination with siTRAP1 showed a reduction of total ATP levels ( Figure 4D ) . Although [A53T]α-Synuclein alone significantly reduced Complex I activity , overall ATP levels were unchanged . Loss of mitochondrial membrane potential predisposes cells to apoptosis . [A53T]α-Synuclein has been suggested to adopt an alpha-helical conformation that could perforate membranes . At the same time , TRAP1 protection against apoptosis has been suggested to act via inhibition of opening mitochondrial permeability transition pore ( PTP ) [49] . The mitochondrial membrane potential is thought to indirectly reflect the state of the PTP . Cells were thus assessed for mitochondrial membrane potential using the mitochondrial membrane dye , JC-1 . Only cells expressing [A53T]α-Synuclein in combination with siTRAP1 showed a loss of mitochondrial membrane potential ( Figure 4E ) . Finally , to exclude the possibility that altered Complex I ATP production might be due to varying quantites of mitochondria within the cells , instead of a functional deficit in the ETC , cell samples were probed for two mitochondrial proteins , VDAC1 and COX4 . No major differences were observed for expression of VDAC1 and COX4 ( Figure S4A ) . This suggests the detected decrease in Complex I ATP production resulted from a functional ETC deficit . JC-1 is an excellent dye to measure mitochondrial membrane potential and because of the color switch following depolarization , it makes it easy to normalize to cell density . However , JC-1 has been superseded by other dyes , like TMRM , with respect to the potential artifact of local concentration changes . With regard to this potential problem , we repeated mitochondrial membrane potential measurements using TMRM . In addition , we wanted to exclude potential off-target effects by siRNA treatment . Therefore , we generated HEK293 cells with stable expression of shTRAP1 constructs resulting in a roughly 90% loss of TRAP1 protein levels ( Figure S5A ) . In stable TRAP1-silenced cells , a significant reduction in membrane potential was observed after [A53T]α-Synuclein expression . This effect was absent in cells expressing scrambled shRNA , again indicating that TRAP1-silencing in combination with [A53T]α-Synuclein expression causes opening of mitochondrial PTP ( Figure S5B ) . The human TRAP1 ATPase domain shares high homology with both other HSP90 proteins and TRAP1 orthologs found in other species ( Figure S6 ) . Recently , the ATPase domain of yeast HSP90 has been shown to be required for its HSP90 function . The mutation of a specific amino acid within the ATPase domain was sufficient to inhibit ATP binding [50] . This amino acid is highly conserved in both HSP90 and TRAP1 proteins ( Figure S6 ) . We therefore exchanged the aspartic acid at position 158 for asparagine ( TRAP1[D158N] ) , creating a putative non-functional ATPase domain . Introducing the D158N mutation did not interfere with TRAP1 protein turnover , as expression in HEK293 cells resulted in similar abundance of TRAP1[D158N] and TRAP1[WT] proteins ( Figure 5A ) . Next , we asked if TRAP1[D158N] is as effective as TRAP1[WT] in protecting [A53T]α-Synuclein-expressing cells from oxidative stress . Cells overexpressing [A53T]α-Synuclein treated overnight with rotenone displayed a robust reduction in cell survival , which was rescued by TRAP1[WT] overexpression ( Figure 4B ) . In contrast , overexpression of TRAP1[D158N] was less effective ( Figure 5B ) . Similar results were observed when we tested ATP production by Complex I . In the context of [A53T]α-Synuclein expression without oxidative stress , TRAP1[WT] rescued [A53T]α-Synuclein-induced decrease in Complex I ATP production ( Figure 5C ) , while TRAP1[D158N] showed significantly lower degree of rescue ability . Finally , cell lysates were again analyzed for abundance of the mitochondrial proteins VDAC1 and COX4 . No changes in VDAC1 or COX4 protein levels were observed in cell lysates expressing either TRAP1[WT] or mutant TRAP1[D158N] ( Figure S4B ) . These data thus indicate that mutant TRAP1 expression does not alter the overall mitochondrial content , arguing in favor of a functional ETC Complex deficit , rather than a deficit due to diminished numbers of mitochondrial/ETC components in the cell . Recent data show that α-Synuclein impairs mitochondrial fusion , leading to fragmented mitochondria . Interestingly , the α-Synuclein-induced mitochondrial fragmentation can be attenuated by co-expression of PINK1 , Parkin and DJ-1 , but not by PD-linked mutant variants of these proteins [51] . Therefore , we sought to determine if TRAP1 is also able of attenuating α-Synuclein-induced mitochondrial fragmentation in SH-SY5Y cells . The [A53T]α-Synuclein-induced punctate mitochondrial staining was reversed to a tubular mitochondrial network by TRAP1[WT] co-expression . In contrast , co-expression of TRAP1[D158N] showed no effect ( Figure 6A , 6B ) . The expression of both TRAP1 variants alone had no impact on mitochondrial integrity . Verification of protein expression levels revealed robust α-Synuclein and TRAP1 expression after transfection with respective plasmids ( Figure 6C ) . Thus , the impaired rescue ability of TRAP1[D158N] in comparison to TRAP1[WT] cannot be explained by the lower abundance of TRAP1[D158N] protein . It is rather the consequence of an altered function of the inherent ATPase function of TRAP1[D158N] . In addition , we asked if reduced TRAP1 levels might enhance mitochondrial fragmentation induced by [A53T]α-Synuclein expression . We noticed that TRAP1-silencing increased the number of cells with fragmented mitochondria . Combining TRAP1-silencing with [A53T]α-Synuclein expression enhanced fragmentation of mitochondria even further ( Figure 7A , 7B ) . Effective TRAP1-silencing and [A53T]α-Synuclein expression was confirmed by Western blot analysis ( Figure 7C ) . TRAP1 is defined as a mitochondrial molecular chaperone and has been shown to be protective against oxidative stress-induced cell death via multiple postulated mechanisms including cytochrome c release inhibition , interference with caspase-3 activation and attenuation of ROS production [39] , [46] , [48] , [52]–[54] . We thus hypothesized that TRAP1 might directly antagonize α-Synuclein mitochondrial-related toxicity . Confirming that TRAP1 is indeed found in the mitochondria , co-localization studies in HEK293 cells revealed a strong overlap between TRAP1 staining with “Mitotracker Orange”-labeled mitochondria ( Figure S7A ) . Therefore , it was interesting to see if α-Synuclein might also localize with mitochondria , as previously reported [33] , [37] , [38] . To determine this , we performed cell fractionation experiments to separate cytoplasmic and mitochondrial enriched fractions . Using Western blotting , these fractions and input control were compared for the content of endogenous , VDAC1 ( mitochondrial outer membrane protein ) , β-Tubulin ( cytosol ) and α-Synuclein proteins . Whereas the input showed abundance of all tested proteins , the cytosolic fraction displayed expected cytosolic proteins β-Tubulin and α-Synuclein . In the mitochondria enriched fraction , no contaminating protein from β-Tubulin could be detected . Importantly , exogenous [A53T]α-Synuclein protein was found within the mitochondria enriched fraction ( Figure S7B ) . Given the strong rescue effect of TRAP1 on toxicity induced by [A53T]α-Synuclein in various systems ( flies , primary rat neurons , and human cells ) , this implies at least a genetic interaction of these proteins . Whether there is a direct interaction of [A53T]α-Synuclein and TRAP1 awaits further analysis .
α-Synuclein plays an important role in PD pathogenesis . However , the mechanisms that actually lead to α-Synuclein-induced neurotoxicity remain unresolved . To gain insights into the disease mechanisms triggered by α-Synuclein , we performed a genome-wide modifier screen on [A53T]α-Synuclein-induced toxicity in flies . We used [A53T]α-Synuclein for our screen because its overexpression in flies results in a robust Parkinsonian phenotype [21] , [55]–[57] . Toxicity induced by α-Synuclein or its mutant variants is rather low and eye-specific expression of A53T does not cause rough eye phenotypes ( REPs ) . Such REPs induced by eye-specific expression of toxic proteins provide an excellent tool for screens and have successfully been used in the past to identify genetic interactions applying alterations in eye morphology due to photoreceptor degeneration as an endpoint . Given the low toxicity of [A53T]α-Synuclein , such screening approaches could not be conducted with regard to α-Synuclein-induced toxicity in flies . Our genetic screen fulfilled two important requirements: it utilized ( i ) an age-dependent model of [A53T]α-Synuclein toxicity , and ( ii ) an endpoint that is relevant to PD , this being the loss of DA . However , apart of being used as a neurotransmitter , DA in flies is also used for cuticle tanning . Thus , we cannot exclude the possibility that cuticle-derived DA might contribute to the overall DA in fly heads . Therefore , the measured decline is not only connected to DA loss in neurons . Nevertheless , secondary readouts like locomotion measurements or DA neuron counts indicate a strong correlation between decreased head DA content and proper function of DA neurons . One of the candidates identified in our screen was the mitochondrial chaperone TRAP1 . Consistent with our results , a genetic screen for alteration of α-Synuclein aggregation , conducted in C . elegans , identified R151 . 7 , a homologue to Drosophila and human TRAP1 , as a candidate worm gene . Knockdown of R151 . 7 resulted in premature α-Synuclein aggregation [58] . Although aggregation was not assayed in our screen , this finding acts as an external confirmation that TRAP1 genetically interacts with α-Synuclein in different in vivo systems . In multiple cell culture systems , TRAP1 has been shown to provide anti-apoptotic functions [48] , [52] , [53] as high levels of TRAP1 reduce the release of key factors involved in apoptosis , including Apoptosis Inducing Factor-1 and Cytochrome c , and additionally prevents Caspase-3 cleavage [39] , [46] , [48] , [59] . The direct mechanisms by which TRAP1 might inhibit apoptosis were not examined in this study . However , given that overexpression of TRAP1 in both rat primary neurons and HEK293 cells was able to enhance cell survival after rotenone treatment , we hypothesize that anti-apoptotic mechanisms might in part be responsible for rescue of [A53T]α-Synuclein toxicity by TRAP1 . This is in agreement with the observation that PD-associated neuronal death involves apoptotic cell death [60]–[63] . In addition , the effects of TRAP1 modulation on ATP synthesis and activities of the ETC support a mitochondrial function . For more than two decades , biochemical studies , the 1-methyl-4-phenyl-1 , 2 , 3 , 6-tetrahydropyridine ( MPTP ) and transgenic animal models have implicated mitochondrial dysfunction in the pathogenesis of PD [5] , [64]–[71] . Genetic data , including mutations in PINK-1 , Parkin , DJ-1 , and HtrA2 , have now specifically linked PD to both dysfunction and morphological change of the mitochondria [72]–[80] . However , the relationship of α-Synuclein pathology and mitochondrial dysfunction has been less clear . Our data are compatible with a localization of [A53T]α-Synuclein either in mitochondria or in mitochondrial membranes . Recent findings , though , have indicated that α-Synuclein may be localized to the outer mitochondrial membrane in pathological conditions and induce morphological changes of mitochondria by inhibiting mitochondrial fusion and enhancing mitochondrial fragmentation [51] . These morphological changes were rescued by overexpression of wild type PINK1 , Parkin , and DJ-1 [51] . We show here that TRAP1 overexpression is also able to reverse [A53T]α-Synuclein-induced mitochondrial fragmentation . Interestingly , TRAP1 has been identified as a substrate of the serine/threonine kinase PINK1 . Phosphorylation of TRAP1 by PINK1 seems to be required for the protective effects mediated by PINK1 . Combining these data with our findings leads to a potential pathogenic model , in which [A53T]α-Synuclein induces mitochondrial stress impairing , most likely , Complex I of the ECT by an as yet unidentified mechanism . Overexpression of TRAP1 counteracts this effect in flies , primary neurons and human neuronal as well as non-neuronal cells . TRAP1[D158N] is less effective in protecting from [A53T]α-Synuclein-induced detrimental effects . The finding suggests that a functional ATPase domain is required for TRAP1 function .
Flies were raised and maintained on standard cornmeal-yeast-molasses-agar food at 25°C unless otherwise noted . Non-RNAi stocks were obtained from the Bloomington Drosophila Stock Centre , UAS-RNAi stocks either from the Vienna Drosophila RNAi Center ( VDRC ) or from National Institute of Genetics ( NIG-fly , Japan ) . Bloomington lines used were: Wild type flies ( Oregon R; referred to as + in text ) , w1118;; P{Ddc-GAL4 . L}4 . 36 ( BL7009; ddc: dopa decarboxylase , aminergic neuron specific driver , referred to in text as ddc-GAL4 ) , w*; P{UAS-lacZ . B}Bg4-2-4b ( BL1777; referred to in text as UAS-LacZ ) , y1w67c23; P{SUPor-P}Trap1KG06242 ( BL14032; referred to in text as TRAP1[KG] ) , w[*]; P{w[+mC] = longGMR-GAL4} ( BL8605; referred to as GMR-GAL4 in text ) , w[1118]; P{w[+mC] = UAS-HsapHSPA1L . W}53 . 1 ( BL7455; expresses HSPA1L , the human homolog of HSP70 under GAL4 control; referred to as HSP70 in text ) , w[1118]; P{w[+mC] = UAS-mitoGFP . AP} ( BL8443; expresses GFP with a mitochondrial import signal; referred to as mito-GFP in text ) and w[*]; P{w[+mC] = UAS-Hsap\MJD . tr-Q78}c211 . 2 ( BL8150; expresses a HA-tagged C-terminal fragment of the human Machado-Joseph Disease/Spinocerebellar Ataxia 3 protein with a 78 repeat polyglutamine tract; referred to in text as PolyQ ) . For stable eye-specific expression of PolyQ , GMR-GAL4 driver was recombined with PolyQ transgene ( GMR>PolyQ in text ) . Fly lines suitable for GFP ( yw;; UAS-GFP ) or human [A53T]α-Synuclein ( yw;; UAS-[A53T]α-Synuclein [41] ) expression ( referred to in text as GFP or A53T , respectively ) . Stable expression under control of the ddc driver ( w[*];; ddc-GAL4>UAS-GFP; w[*];; ddc-GAL4>UAS-GFP ) were generated by recombination ( flies referred to in text as ddc>GFP or ddc>A53T ) . Transgenic flies expressing human TRAP1 ( hTRAP1 ) were generated by BestGene Inc . In brief , human TRAP1 cDNA was sub-cloned from pcDNA3 . 1+ vector into pUASattB using KpnI and XbaI restriction sites . hTRAP1 expression in these transgenic flies ( w;; UAS-hTRAP1/TM3 , Sb; referred to in text as hTRAP1 ) was verified by Western blotting . The Bloomington Deficiency Kit was utilized for screening purposes ( http://flystocks . bio . indiana . edu/Browse/df/dfkit_retired_July2009 . htm ) . In general , a specific deficiency line was crossbred with ddc>A53T flies . Male offspring ( ddc>A53T flies in combination with the respective deficiency ) was selected and aged . At ages 1 and 4 weeks , a minimum of 9 flies was collected for later measurement of head DA using HPLC . Liquid nitrogen flash frozen fly heads were homogenized ( Precellys 24 homogenizer ) in homogenization buffer ( 0 . 1 M perchloric acid/3% trichloric acid solution ) . 50 µl of supernatant from each sample were used for HPLC analysis ( Dionex Ultimate 3000; running buffer: 57 mM citric acid , 43 mM sodium acetate , 0 . 1 mM EDTA , 1 mM octane sulfonic acid , 20% methanol ) . Samples were separated on a chromatographic column ( Dionex Acclaim C18 , 5 µm , 2 . 1×150 mm column , at 25°C ) , and DA was electrochemically detected on a graphite electrode ( Dionex ED50 Electrochemical detector with following conditions: disposable carbon electrode at 0 . 8 V , flow rate 0 . 2 ml/min ) . DA ( Sigma-Aldrich ) standards of 0 . 1 µM , 0 . 25 µM and 0 . 4 µM were used for creation of a standard curve . Chromeleon 6 . 6 software was used for HPLC data analysis . Longevity assays were performed as previously described [81] . For oxidative stress assays a minimum of 20 male flies ( 2–3 days of age ) was kept on filter papers soaked with paraquat ( 20 mM paraquat dichloride in 5% sucrose ) . Survival of flies was scored on a daily base . Fresh paraquat/sucrose solution was supplied daily . Fly climbing was assessed in accordance with previously published protocols [21] , . Flies were aged on normal yeast medium . At ages 1 and 4 weeks , climbing was assessed ( 20 flies per genotype ) . Flies were individually placed in a graduated cylinder , and allowed to climb for 15 s . Maximum height attained was recorded , and analysis was repeated 3 times per time point , with 3 trials at one minute intervals recorded at each time point . Fly brains were dissected in cold PBS , washed in a PBS/0 . 1% Triton X ( PBT ) , fixed in 4% PFA ( 30 min , 4°C ) , and blocked in PBT containing 5% normal goat serum ( overnight , 4°C ) . For TH staining , brains were incubated with primary anti-TH antibody ( 1∶100; rabbit polyclonal , AB152 , Chemicon International/Millipore ) for 2 days , 4°C , and subsequently with fluorescent secondary anti-rabbit antibody ( 1∶200; AlexaFluor-555 or Cy3; Invitrogen/Jackson Immunological Research ) for 3 . 5 hours . Afterwards , brains were mounted in Vectashield ( Vector Labs ) . The number of TH-positive neurons was determined on Z-stacked confocal sections ( 1 µm , Leica DM IRE2 , Laser ) [83] . At least 15 brains were analyzed per genotype . Fly heads were homogenized in RIPA buffer ( 50 mM Tris , pH 8 . 0 , 0 . 15 M NaCl , 0 . 1% SDS , 1 . 0% NP-40 , 0 . 5% Na-Deoxycholate , 2 mM EDTA , Complete Protease Inhibitors ( Roche Applied Sciences ) , pH 7 . 4 ) , centrifuged , and the supernatant was collected . Cell culture protein samples were collected after washing cells in ice cold PBS , followed by lysis in RIPA buffer for 30 min on ice . Cell debris was removed by centrifugation , and supernatants were collected . For Western blot analysis , protein samples were separated via SDS-PAGE gel and then transferred onto nitrocellulose membrane . Blocking in skim milk was followed by overnight primary antibody incubation . The primary antibodies used were as follows: mouse anti-α-Synuclein ( 1∶1000; Cell Signaling ) ; mouse anti-Drosophila Syntaxin ( 1∶2000 Developmental Studies Hybridoma Bank ( DSHB ) ) , mouse anti-β-Tubulin ( 1∶500 , DSHB ) ; mouse anti-Δ Tubulin ( 1∶10 , 000; Sigma-Aldrich ) ; mouse anti-TRAP1 ( 1∶1000; BD Biosystems ) ; mouse anti-phospho-tyrosine ( PY99 ) ( 1∶200; Santa Cruz Biotechnology ) ; mouse anti-phospho-threonine ( H2 ) ( 1∶200: Santa Cruz Biotechnology ) ; mouse anti-phospho-serine ( 16B4 ) ( 1∶200; Santa Cruz Biotechnology ) ; mouse anti-Cytochrome c ( 1∶500; Santa Cruz Biotechnology ) ; rabbit anti-VDAC1 ( 0 . 3 µg/ml; Abcam ) ; mouse anti-COX IV ( 2 µg/ml; Abcam ) ; rabbit anti-GFP polyclonal ( 1∶1000; Santa Cruz Biotechnology ) . Appropriate secondary anti-mouse or rabbit horseradish peroxidase-linked antibodies ( 1∶10 , 000 ) were obtained from GE Healthcare . Membranes were incubated with the secondary antibody for one hour , followed by signal detection using the Chemiglow substrate ( Biozym ) . Method for fly head RNA isolation was adapted from the following link: http://www . ou . edu/journals/dis/DIS84/Tec2%20Bertucci/Bertucci . htm . 20 fly heads per tube were used for RNA isolation . RNA samples were treated with DNase following manufacturer's instructions ( Promega RQ1 RNase-Free DNase kit ) . Total RNA from cultured cells was prepared from cells using Qiagen RNeasy Mini kit ( Qiagen ) . RNA was used for cDNA production via reverse transcription using the iScript cDNA Synthesis Kit ( BioRad ) . Real-time PCR measurements were performed using the SYBR Green ( Thermo Fisher Scientific ) reagent following manufacturer's instructions for preparation of PCR samples . Gene of interest signal was compared to that of control gene expression ( β-Actin5c for fly samples and 18S for human samples ) using the 2−ΔΔCt method [84] . No-RT controls were performed to exclude for genomic DNA sample contamination . PCR reactions were followed by generation of a dissociation curve to check for side product generation . Amplification conditions for fly samples were as follows: 5 min at 95°C , 40 cycles of: 30 s at 95°C , 30 s at 58°C , 60 s at 72°C , followed by 10 min at 72°C . Gene of interest was normalized to control β-Actin5c signal . The primers used are listed in Table S5 . Amplification conditions for cell culture samples were as follows: 5 min at 95°C , 40 cycles of: 30 s at 94°C , 30 s at 60°C , 60 s at 72°C , followed by 10 min at 72°C . Gene of interest was normalized to control 18S signal . For primers , see Table S5 . Cells were incubated for 16 hours in the presence of either hydrogen peroxide ( 100 µM ) or rotenone in DMSO ( HEK293: 200 µM , rat cortical: 1 µM rotenone ) . Rotenone control cells were treated with equivalent amount of DMSO alone . After overnight oxidative stress treatment , cells were fixed in 4% PFA for 10 min , before permeabilization in PBT for 10 min . Cells were incubated with Hoechst nuclear stain for 30 min before imaging on a fluorescent microscope ( Leica DMI6000B ) . Using a macro within the Leica Qwin V3 quantification software , cell number remaining in each well was assessed by counting total fluorescent nuclei ( 6 images per well at 10× , with minimum 6 wells per genotype in a 24-well culture dish ) . Method was adapted from [86] . Cells were trypsinized and washed three times in cold PBS . Cells were then resuspended in incubation medium ( 2×105 cells/ml , 25 mM Tris , 125 mM KCl , 2 mM K+EDTA , 10 mM KH2PO4 , pH 7 . 4 ) and permeabilized with digitonin ( 40 µg/ml ) . To measure ATP produced via complexes I , III , IV , 2×104 cells ( minimum of 8 replicates per experiment ) were resuspended in a complex incubation medium supplemented with 1 mg/ml BSA , 2 mM ADP , 10 mM glutamate , 2 mM malate . After incubation for 20 min at 37°C , the reaction was stopped by addition of 6% perchloric acid , and samples were neutralized with 3 M K2CO3 . ATP in each sample was measured in a plate reader using the CellTiter-Glo Luminescent reagent ( Promega ) following manufacturer-provided instructions . Total ATP levels were obtained using an ATP standard curve ( ATP from Sigma ) , with final calculation expressing ATP as pmoles ATP per minute per 1 million cells , and plotted as percentage of control values . 2×104 HEK293 cells/well were seeded ( 12 replicates per experiment ) in an opaque white 96-well plate . Total ATP levels were measured as stated above . Following manufacturer-provided instructions , cells ( 6 replicates per experiment ) in a 96-well plate ( black sided , clear bottom ) were incubated with the mitochondrial probe JC-1 ( 3 µg/ml; Invitrogen ) in full medium for 30 min at 37°C . After washing in PBS , JC-1 mitochondrial aggregates were measured in a plate reader ( excitation: 530 nm , emission: 590 nm ) . As a control , JC-1 fluorescence was measured in the presence of the mitochondrial membrane potential inhibitor CCCP ( Carbonyl cyanide m-chlorophenylhydrazone , 50 µM ) . Protein content per well was quantified using the Bradford assay . Fluorescence in relation to total protein was displayed as percentage of control values . Cells were plated on PLL-coated glass slips . 48 hours after transfection , cells were fixed in 4% PFA for 10 min , permeabilized in PBT for 10 min , followed by blocking in 1% BSA and overnight incubation with the primary antibody at 4°C . Antibodies used: monoclonal rat anti-α-Synuclein ( 1∶500 , Alexis Biochemicals/Enzo Life Sciences ) ; monoclonal mouse anti-NeuN ( 1∶500 , Chemicon ) ; monoclonal mouse anti-TRAP1 ( 1∶300 , Alexis Biochemicals ) . For visualization of mitochondria cells pretreated for 4 hours with 1 µM rotenone , the cells were incubated with Mitotracker Orange CMTMRos ( 300 nM , following manufacturer's instructions , Invitrogen ) for 30 min at 37°C prior to fixation . Cells were incubated with respective secondary antibodies for one hour ( all secondary antibodies 1∶1000 , anti-mouse or rat AlexaFluor-488 , 543 , 633 , Invitrogen ) , and mounted using Mowiol ( Calbiochem ) , with or without the anti-bleaching agent DABCO or nuclear stain Hoechst ( Sigma-Aldrich ) . Mitochondria were isolated from HEK293 cells transfected with [A53T]α-Synuclein using the following protocol: Cells were suspended in MB buffer ( 70 mM sucrose , 10 mM HEPES , 1 mM EDTA , 210 nM Mannitol ( pH 7 . 5 ) and protease inhibitors ) , homogenized with an injection needle ( 27G ¾″ 19 mm , 5–6 strokes ) and centrifuged at 750×g for 7 min . After centrifugation the pellet was resuspended in MB buffer , homogenized using the same injection needle and centrifuged . This procedure was repeated four times . The resulting supernatants were pooled and centrifuged at 10000×g for 30 min . This mitochondria-containing pellet was resuspended in MB buffer and further centrifugated at 1500×g for 20 min . The purity of the resulting mitochondrial pellet was examined by Western blotting using specific antibodies directed against β-Tubulin , VDAC and α-Synuclein . Data was analyzed using GraphPad Prism 4 . 0 ( GraphPad Software Inc . ) , using 1-way ANOVA followed by Newman-Keuls post testing . Use of a 2-way ANOVA was noted in the text . Survival data were analyzed with the Kaplan-Meier analysis method and the Log Rank Test for curve statistical comparison analysis . Statistical significance referred to as: *p<0 . 05; **p<0 . 01; ***p<0 . 001 . All data is presented as mean ± SEM . | Parkinson's disease ( PD ) is a progressive neurodegenerative disorder , pathologically characterized by loss of dopaminergic neurons in the substantia nigra pars compacta brain region . Mutations in α-Synuclein or gene duplication or triplication result in autosomal-dominant inherited PD . Indeed , aggregated and insoluble α-Synuclein is found in Lewy bodies , a pathological hallmark common to both sporadic and hereditary forms of PD . In order to better define α-Synuclein's pathogenic mechanism , we first used a fly genetic screen to search for novel genetic modifiers of mutant human [A53T]α-Synuclein neurotoxicity . We identified the mitochondrial chaperone protein TRAP1 as a novel modifier of the toxicity induced by [A53T]α-Synuclein . [A53T]α-Synuclein–induced toxicity was enhanced when TRAP1 expression was decreased , while overexpression of human TRAP1 ( hTRAP1 ) provided a rescue . Cell culture experiments further demonstrated that [A53T]α-Synuclein directly interferes with a number of mitochondrial functions , including Complex I ATP production , mitochondrial fragmentation , and sensitivity to oxidative stress . These effects could be blocked by TRAP1 overexpression . As mitochondrial dysfunction has previously been linked to mutations in several other genes associated with genetic PD , these data provide further evidence of a common mitochondrial-centric mechanism of PD pathogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"animal",
"models",
"model",
"organisms",
"genetics",
"biology",
"neuroscience",
"genetics",
"and",
"genomics"
] | 2012 | The Mitochondrial Chaperone Protein TRAP1 Mitigates α-Synuclein Toxicity |
While circadian dysfunction and neurodegeneration are correlated , the mechanism for this is not understood . It is not known if age-dependent circadian dysfunction leads to neurodegeneration or vice-versa , and the proteins that mediate the effect remain unidentified . Here , we show that the knock-down of a regulator ( spag ) of the circadian kinase Dbt in circadian cells lowers Dbt levels abnormally , lengthens circadian rhythms and causes expression of activated initiator caspase ( Dronc ) in the optic lobes during the middle of the day or after light pulses at night . Likewise , reduced Dbt activity lengthens circadian period and causes expression of activated Dronc , and a loss-of-function mutation in Clk also leads to expression of activated Dronc in a light-dependent manner . Genetic epistasis experiments place Dbt downstream of Spag in the pathway , and Spag-dependent reductions of Dbt are shown to require the proteasome . Importantly , activated Dronc expression due to reduced Spag or Dbt activity occurs in cells that do not express the spag RNAi or dominant negative Dbt and requires PDF neuropeptide signaling from the same neurons that support behavioral rhythms . Furthermore , reduction of Dbt or Spag activity leads to Dronc-dependent Drosophila Tau cleavage and enhanced neurodegeneration produced by human Tau in a fly eye model for tauopathy . Aging flies with lowered Dbt or Spag function show markers of cell death as well as behavioral deficits and shortened lifespans , and even old wild type flies exhibit Dbt modification and activated caspase at particular times of day . These results suggest that Dbt suppresses expression of activated Dronc to prevent Tau cleavage , and that the circadian clock defects confer sensitivity to expression of activated Dronc in response to prolonged light . They establish a link between the circadian clock factors , light , cell death pathways and Tau toxicity , potentially via dysregulation of circadian neuronal remodeling in the optic lobes .
Alzheimer’s disease ( AD ) is a neurodegenerative disorder that involves neuronal cell loss , extracellular amyloid plaques , and intracellular neurofibrillary tangles . During AD and other neurodegenerative diseases , neurons induce a series of proteases , including caspases , and a number of key proteins are cleaved by caspases including APP , Presenilin ( PS1 , PS2 ) , Tau and Huntingtin [1–5] . This has led to the suggestion that the extensive neuronal loss observed in AD may result from the activation of apoptotic related pathways [6] . Caspases can be activated within a cell without immediately causing classical apoptosis , and there is evidence for prolonged caspase activation without neuronal death [7] . For instance , in AD chronic caspase activation may lead to cleavage of Tau and other essential cellular proteins and contribute to neuronal pathology prior to cell death [3] . Caspase-cleaved Tau is present in AD , but not control brain [8 , 9] . Additionally , caspase-cleaved Tau is more fibrillogenic in vitro than full-length Tau [9] . Drosophila melanogaster has been used as a model system for the analysis of AD [10] . The Drosophila Tau protein is similar to the mammalian one and accumulates in axonal processes [11] . When it is overexpressed in neurons or eyes , fly Tau induces apoptotic neuronal cell death [12] . In addition , synaptic dysfunctions induced by human or fly Tau have been produced and analyzed in Drosophila larval motor neurons and neuromuscular junctions . Expression of Tau within motor neurons generates altered morphology in the presynaptic terminals and defective synaptic transmission and microtubule-based axonal transport [13 , 14] . The fly mushroom body is the key locus for olfactory learning and memory in Drosophila , where human or fly Tau expression causes an impairment of associative learning and memory followed by neurodegeneration [15] . Finally , when expressed in the fly eye or brain , human Tau produces aspects of human tauopathies , including neurodegeneration [16 , 17] . Like many organisms , Drosophila operates on a 24-hour cycle that is maintained by environmental input to an internal body clock [18] . The clock depends on oscillations in the activation of specific genes at certain times of the day . The key feature of these oscillations is a negative feedback loop , in which transcriptional regulators like Period ( Per ) repress transcription of their mRNAs . The Drosophila Doubletime ( Dbt ) protein is homologous to mammalian Casein Kinase 1 and phosphorylates Per monomers , resulting in Per degradation [19 , 20] . In addition to its role in regulating circadian rhythms , Dbt has also been shown to be involved in regulating cell death pathways . For example , overexpression of Dbt in the fly eye has been shown to rescue the eye morphology defect caused by expression of proapoptotic proteins such as Reaper and Hid [21] . Circadian rhythm disturbances affect as many as 25% of AD patients during some stage of their disease [22] . As a consequence , sleep , the biological clock , and core body temperature are affected . Some common symptoms of AD that are related to disturbances in the circadian clock are insomnia , nocturnal behavioral changes , and excessive daytime sleepiness . A postmortem study found significant differences in the expression pattern of circadian genes between Alzheimer patients and controls [23] . In addition , the 3xTg ( triple transgenic ) mouse models of AD , which exhibit Tau neuropathology , showed deteriorated circadian organization of locomotor behavior [24] . However , it is not known if circadian components are directly linked to disease onset or if circadian dysfunction is just a consequence of AD . Our search for proteins that interact with the Drosophila circadian kinase Dbt has led to results demonstrating that Dbt and one of these interactors connect the circadian , cell death and neurodegenerative pathways . Initially , a screen was performed for effectors of Dbt’s circadian function; we screened a list of candidate genes ( mostly phosphatase catalytic subunits ) with dsRNAi lines crossed to the timGAL4 driver for those that would alter circadian periods and lead to changes in Dbt electrophoretic mobility , potentially indicative of those that would lead to autophosphorylation of Dbt . This screen identified spaghetti ( CG13570 , or spag ) . spag encodes a tetratricopeptide repeat ( TPR ) -containing protein initially identified in a screen for modifiers of protein aggregation in Huntington disease and was reported to interact both with Huntingtin protein and the chaperone protein HSP90 [25] . Here , we show that spag knock-down or reduced Dbt activity in circadian cells leads to longer circadian periods and expression of activated initiator caspase only during the middle of the day . Links to the circadian clock are shown by the effects of light , circadian mutants and circadian cells , while links to AD are shown by the resulting cleavage of Tau , enhancement of neurodegeneration , reduced healthspan and effects of aging .
We targeted the clock cells in flies with spag RNAi knock-down ( timGAL4>UAS-dcr; UAS-spag RNAi ) and observed altered locomotor activity rhythms . spag knockdown flies exhibited long periods ( 25 . 5–26 . 5 h ) or arrhythmic locomotor activity ( S1 Fig and Table 1 ) , depending on the specific RNAi transgene , the inclusion of dcr in the genotype and the duration of transgene expression after eclosion . Null mutations of spag and ubiquitous knock-down of spag were lethal , so all of our analysis has employed knock-downs of spag in circadian cells with the timGAL4 or pdfGAL4 drivers . In addition to locomotor defects , knockdown of spag led to a decrease in Dbt protein levels and an increase in the levels of activated initiator caspase Dronc , detected with an antibody that only detects the cleaved ( and thereby activated ) form of Dronc [26] ( Figs 1A and S2A; See Materials and Methods ) . Interestingly , the decrease in Dbt and the accumulation of activated Dronc occurred mostly during the daytime close to ZT7 ( Lights are on from ZT0-12 and off from ZT12-24 ) . To determine the time course more precisely , head extracts were collected at one hour intervals from ZT1-7 , and Dbt disappearance and activated caspase were detected by ZT6 ( Fig 1B ) . In addition , the levels of activated Dronc were reduced after ZT9 and gone by ZT11 ( Fig 1B ) . These effects of spag knock-down were further confirmed in Drosophila S2 cells , which lack a circadian rhythm , and knockdown of spag with two different non-overlapping dsRNAi’s led to a decrease in Dbt levels , followed by accumulation of activated Dronc ( Figs 1C and S3A ) . The loss of Dbt from Drosophila S2 cells or fly heads was preceded by a post-translational modification of Dbt that produced a mobility shift with SDS-PAGE ( S2B–S2E Fig ) , and in fly heads the loss was not always complete but was enhanced at higher temperatures , longer times post-eclosion and by light rather than darkness ( S2C , S2D , and S2E Fig ) . Light is a strong trigger , as Dronc was activated 7 hrs after light periods starting at ZT13 in a number of lines ( including spag RNAi ) that produced activated Dronc during the day ( S2F Fig ) . A variable amount of Dbt reduction was also observed in the middle of the night ( ZT19; Figs 1A and S3B ) , although activated Dronc was not detected . In summary , reduced spag levels in circadian cells triggers post-translational modification of Dbt , reduced Dbt levels and accumulation of activated caspase at specific times of day . To address whether the caspase activation was a consequence of loss of Dbt function we expressed the kinase-dead ( DbtK/R ) form of Dbt in circadian cells with a timGAL4 driver . The DbtK/R protein acts as a dominant negative to antagonize endogenous Dbt [27] . DbtK/R flies also showed elevated caspase activity at ZT7 , while flies expressing DbtWT did not ( Fig 1D ) . The results suggest that the kinase activity of Dbt negatively regulates expression of activated Dronc , and that reductions in this activity can cause expression of activated Dronc . To determine if other clock mutants showed activated caspase expression we collected heads of per0 and ClkJrk mutant flies and analyzed for activated caspase expression . ClkJrk flies showed activated caspase expression at ZT7 like flies expressing DbtK/R , but per0 did not ( Fig 1E ) , and more extensive analysis of ClkJrk flies showed activated caspase only at ZT7-9 in ClkJrk flies ( S4 Fig ) . timGAL4>UAS-dbtK/R flies express high levels of PER [27] , which should repress the CLK/CYC transcription factor and thereby produce a condition like that found in ClkJrk flies , which lack CLK-dependent transcription . Therefore , activated caspase is produced in two different circadian mutants with similar effects on the circadian transcription cycle . In order to determine if the time of the circadian clock ( e . g . , ZT7 ) is needed for production of activated Dronc , we examined the timing of Dronc induction in perS ( or perL ) ; timGAL4>UAS-dbtK/R flies , along with its activation in ClkJrk flies and wild type flies , and in all mutant flies activation was detected from ZT7-9 ( S4 Fig ) , despite the fact that the perS mutation significantly shortened the period of locomotor activity for the timGAL4>UAS-dbtK/R genotype ( 31 . 0 ± 0 . 4 , n = 14 vs 33 . 7 ± 0 . 9 , n = 15 for the perS and per+; dbtK/R genotypes respectively; all perL; dbtK/R flies , n = 16 , were arrhythmic ) . Taken together with the ZT7 time of Dronc activation for the largely arrhythmic timGAL4>UAS-dbtK/R and ClkJrk genotypes , the absence of Dbt reductions in DD ( S2E Fig ) , and the production of activated Dronc in light-pulsed flies at night ( S2F Fig ) , these results suggest that the production of activated caspase is a transient response to prolonged light exposure that also involves reductions in several circadian genes ( e . g . , spag , dbt and clk ) . S2 cells were treated with spag dsRNA with and without the proteasome inhibitor MG132 and immunoblotted for Dbt . Endogenous Dbt levels were stabilized in the presence of the proteasome inhibitor in the presence of spag RNAi , which led to complete absence of Dbt in the absence of proteasome inhibitor , but higher Dbt levels in the presence of proteasome inhibitor and spag RNAi were obtained than with proteasome inhibitor only ( Fig 1F ) . The higher levels of Dbt in the presence of spag RNAi and proteasome inhibitor than with proteasome inhibitor only , coupled with the lower levels of Dbt with spag RNAi only , suggest that Spag may have both proteasome-dependent positive and proteasome-independent negative effects on Dbt levels . This result also demonstrates that Dbt is degraded by the proteasome in response to spag knock-down . The transient accumulation of forms of Dbt with slow electrophoretic mobility ( S2B–S2E Fig ) suggests that phosphorylation and ubiquitination of Dbt is the initial consequence of the spag knock-down ( See [28 , 29] for evidence of phosphorylation . ) , with proteasomal degradation of Dbt a subsequent consequence . When DbtWT is expressed in fly head circadian cells in a spag RNAi background accumulation of activated Dronc is blocked ( Fig 2A ) , suggesting that the effects of spag knock-down are mediated by reductions in Dbt . In addition , when Dbt is overexpressed in S2 cells it is able to block the accumulation of activated caspase associated with spag or dbt RNAi ( Fig 2B ) . However , Spag overexpression was not able to block activated caspase accumulation brought on by dbt RNAi but did block activated caspase accumulation by spag RNAi , suggesting that Spag is upstream of Dbt and confirming the specificity of the dsRNAi knock-down for spag ( Fig 2B ) . Fly brains from timGAL4>UAS-spag RNAi flies collected at ZT7 showed elevated levels of active caspase not found in control brains , but not at ZT19 ( Fig 3A and 3B ) . This was also observed using the pdfGAL4 driver , which is expressed in the PDF-secreting brain neurons that drive circadian rhythms of locomotor activity in constant darkness , and with expression of the dominant negative DbtK/R with both drivers ( Figs 3D and 4A for ZT7 and S5A for ZT19 ) . tim-GAL4 clock cells expressing DbtK/R expressed both activated caspase and Dbt-MYC signal , but activated caspase was expressed in other cells as well ( Fig 3D ) . Taken together , these results suggest both autonomous and non-autonomous effects of the transgene expression . The elevated activated Dronc was particularly prominent in the optic lobe where the PDF+ axons terminate , but it was found in other brain-associated tissues as well . In addition , Pdf receptor mutant brains lacked this caspase activation that was observed at ZT7 ( Figs 4B for ZT7 and S5B for ZT19 ) . Moreover , most of the activated caspase was detected in tissue surrounding the PDF+ axons in the optic lobes ( S5C Fig ) . Taken together with the generation of activated caspase in large areas of the brain in pdfGAL4>UAS-spagRNAi ( or—dbtK/R ) flies , the results demonstrate that signaling by PDF is required for this broad activation pattern , and in fact it is even needed in an autocrine manner for expression in the PDF+ cells that also express PDF receptor [30] , because no activation is seen in these cells in the absence of the PDF receptor ( Fig 4B ) . The CM1 antibody , a marker for Caspase-9 like Dronc activity , was also used to detect Dronc activity , and showed a similar pattern as the anti-Dronc antibody used ( Fig 3C ) . In addition , RNAi to Dronc eliminated most of the signal , confirming that the caspase signal we detect is indeed from Dronc ( Fig 3C ) . Since all Dronc signal is eliminated with expression of dronc RNAi with the timGAL4 driver , the results suggest that broader Dronc activation requires Dronc activation in TIM+ cells , which then signal a corresponding increase in the surrounding tissue via PDF signaling . Since spag was initially found in a screen involving neurodegeneration and caspases have been implicated in the cleavage of key proteins associated with diseases such as AD , we examined whether Tau was a substrate for Dronc . First , we expressed HA-tagged Drosophila Tau ( dTau ) in S2 cells and either treated the cells with UV irradiation to induce widespread caspase activation or used RNAi to knock down either Spag or Dbt . When either Spag or Dbt was knocked down cleavage of dTau-HA was detected ( Fig 5A ) . To confirm that Dronc was the main caspase involved in this cleavage we used RNAi and targeted either Dronc or Drice , the other main caspase involved in cell death . When Dronc was silenced along with Spag or Dbt , cleavage was inhibited , but when we targeted Drice the cleavage was still detected ( Fig 5A ) . To further confirm that Dronc was targeting dTau we incubated the dTau lysate with active recombinant Dronc . Samples from cells that were UV-irradiated showed a cleavage product for dTau . In addition , when active Dronc was used we also detected the same cleavage product ( Fig 5B ) , which was not detected with lysates of S2 cells not treated with UV or active Dronc . This suggests that dTau is indeed a substrate for Dronc . Expression of human Tau ( hTau ) in fly eyes produces neurodegeneration and has been used as a fly model for tauopathies [16 , 17] . Therefore , we expressed DbtWT or DbtK/R along with hTau in the fly eye using the eye specific GMR driver to determine if Dbt might enhance eye neurodegeneration . When DbtK/R was expressed activated Dronc was detected , both with and without hTau expression , while expression of hTau alone was not sufficient to activate Dronc ( Fig 5C ) . Expression of DBTK/R and hTau with GMR-GAL4 led to significant reductions in Tau levels at ZT7 ( Fig 5C ) . Moreover , when DbtK/R was expressed along with hTau there was significantly increased disruption in the eye ( Fig 5D and S1 Table ) . The enhanced disruption of the eye was manifested by the appearance of melanized patches in the eye ( Fig 5D ) and a decreased average surface area of the eye ( S1 Table ) . Expression of DbtWT did not lead to enhancement of the eye phenotype ( Fig 5D and S1 Table ) . To determine whether Dronc activation led to classical apoptosis we looked at another marker of apoptosis . Diap1 is an antiapoptotic protein that regulates cell death in flies by binding to and inhibiting the activity of Dronc . When cell death occurs Diap1 is targeted for degradation/cleavage , thereby freeing Dronc [31] . We examined Diap1 levels in fly heads and observed no difference between young wild type and timGAL4>UAS-spag RNAi or dbtK/R flies in which Dronc activation occurs ( Fig 6A ) . This lack of effect on Diap1 levels was also observed in S2 cells treated with spag or dbt RNAi or expressing DbtK/R ( Fig 6D ) . However , when flies aged a Diap1 cleavage product was detected in the spag RNAi and dbtK/R flies ( Fig 6B ) and activated caspase expression was also detected . Interestingly , aged spag RNAi flies showed activated caspase expression at all time points examined ( Fig 6B ) . To determine if this pathway occurred naturally in aging flies we collected wild type Canton S ( C . S . ) fly heads at 60 days . Aged flies showed a mobility shift of Dbt at ZT 7 and 19 that was similar to what we observed in spag RNAi flies ( Fig 6C ) . In addition , increased active caspase expression was detected at ZT7 and 19 for the aged flies ( Fig 6C ) and Diap1 cleavage was detected from ZT7 to 19 for the aged W . T . flies ( Fig 6C ) . These results demonstrate that wild type flies exhibit the same circadian-dependent activation of apoptotic pathways that are produced in spag RNAi , dbtK/R and ClkJrk flies at younger ages , and suggest that reductions in activity of these circadian genes accelerate an age-dependent pathway that leads to activation of the apoptotic pathway . Since spag regulates a pathway that leads to the activation of caspases by reduction of Dbt , and this then leads to the cleavage of Tau by these activated caspases and ultimately to expression of apoptotic markers , we wanted to determine if there were behavioral and lifespan manifestations . The circadian locomotor assays initially suggested that there were effects as flies age . Locomotor behavior is preserved in young flies . However , as transgenic flies with reduction of spag by RNAi age , they tend to die during the locomotor assay ( Table 1 ) . The loss of the climbing response has been used to monitor age-related changes in Drosophila . Normal Drosophila show a strong negative geotactic response . When tapped to the bottom of a vial they rapidly climb to the top of the vial , and most flies remain there . Flies with knockdown of spag initially climb as well as control flies . However , over time they decline in performance more rapidly than controls ( Fig 7A ) . The progressive , accelerated decline in climbing ability in spag RNAi flies demonstrates a functional deficit produced by knockdown of spag in clock cells . In addition , spag RNAi flies had a reduced lifespan compared to control flies ( Fig 7B ) . By contrast , expression of DbtK/R in circadian cells did not produce accelerated death and loss of climbing ability in comparison with wild type Canton S flies , indicating that the effects of Dbt activity reduction are not as severe as those of spag reduction ( See figure legend for discussion of statistics ) .
We have identified a new player ( spag ) that links circadian signaling , cell death and tauopathies together . Orthologs of Spag in yeast and humans function as co-chaperones of Hsp90 to regulate its activity and recruit client proteins for Hsp90 . They are part of multiprotein complexes that contribute to biogenesis of cellular machineries like RNA polymerase , ribonucleoproteins and Phosphatidyl-Inositol 3-kinase-related kinases [32 , 33] . Recently , Drosophila Spag has been shown to associate with Hsp90 and Hsp70 to likewise contribute to assembly of several of these factors [34] . The human ortholog of Spag ( RPAP3 ) is a binding partner for a WD40 repeat protein that is involved in apoptosis . RPAP3 contains TPR domains and regulates apoptosis induced by several stimuli [35 , 36] . When we knock down spag in circadian cells using RNAi we observed reduction in Dbt levels and an increase in the activated caspase Dronc in fly heads . Interestingly , this mostly occurred during the day ( ZT7 ) or after extended light treatment at night , increased as flies age and did not occur in constant darkness . Accumulation of activated Dronc was also observed with expression of the kinase dead form of Dbt ( DbtK/R ) in circadian cells . This suggests that some factor present or active during the light might regulate the Spag-Dbt pathway and confer a transient sensitivity to caspase activation after extended light treatment . One feature that is common to various neurodegenerative diseases is the acceleration of the age-related disruption of the daily cycle of sleep and wake . Our work suggests that the most immediate way for the clock to influence neurodegeneration is by circadian gene-dependent control over the expression of pro-neurodegenerative factors . While these factors are not expressed in young flies with normal circadian clocks , the mutant clocks that we have produced in flies resemble those produced in aging wild type flies , in which Dbt modification , activated caspase expression and cleaved Diap1 are detected at ZT7 and ZT19 . Prior work has shown the circadian function is blunted along with reduced healthspan in aging flies [37–39] . Circadian dysfunction would enhance their susceptibility to light-dependent neurodegeneration . In a previous report , the period gene was mutated in flies with sniffer gene mutations causing neurodegeneration . The flies with both mutations displayed faster neurodegeneration and had shorter lifespans compared to flies with single mutations . This suggests that disrupted circadian rhythms can accelerate the process of neurodegeneration [40] . What is the nature of this mechanism ? Previous work linked Spag to Huntington’s disease and Hsp90 and a possible role in aggregation [25] . In addition , caspases have been shown to be involved in the cleavage of Tau , a protein associated with AD , and this caspase-mediated cleavage of Tau is associated with its aggregation . This led us to examine whether Tau was cleaved by the caspase Dronc in the Spag-Dbt pathway . HA-dTau was cleaved in S2 cells when we knocked down either Dbt or Spag , and this cleavage was prevented when we knocked down Dronc , but not the effector caspase Drice . This result is consistent with cleavage of Tau by Dronc in response to lowered Dbt and Spag activity . We examined whether long-term reductions of Spag and Dbt activity in circadian cells have any detrimental effects on the fly . While younger flies with reduced Spag or Dbt activity did not exhibit expression of cell death marker Diap1 , older flies with chronic reductions in Spag or Dbt activity exhibited elevated levels of Diap1 cleavage—a marker of the apoptotic pathway . These results suggest that chronic reductions in Spag or Dbt activity eventually produce deleterious effects . Flies with reduced spag levels had more cleaved Diap1 compared to dbtK/R flies , higher levels of active Dronc and more rapid decline of climbing proficiency and lifespan . This puzzled us since both lines activate Dronc and lead to dTau cleavage , so we expected identical phenotypes . One possibility is that since Spag has been shown to interact with Hsp90 , Spag might regulate Hsp90 and control the level of aggregation that occurs . In such a model , if Spag is eliminated and can no longer regulate Hsp90 , Hsp90 might no longer interact with dTau and therefore no longer function to reduce dTau aggregates . The defect in Hsp90 function is not predicted for dbtK/R flies , which retain Spag . In addition , Hsp90 is a known regulator of cell death and has been shown to inhibit apoptosis [41] . Removal of Spag might lead to dysregulation of Hsp90 , preventing it from regulating components of the cell death pathway and causing a higher level of activated caspase than observed with Dbt inhibition alone . Is this pathway evolutionarily conserved ? To address this we used the fly eye and expressed human Tau along with DbtWT or DbtK/R . When hTau and DbtK/R were coexpressed together an enhanced disrupted eye phenotype was produced together with Donc activation and Tau cleavage , and the phenotype is less severe when hTau is coexpressed with DbtWT . Since loss of Dbt kinase activity leads to Dronc activation , Dbt may be inhibiting Dronc by direct phosphorylation of Dronc , or alternatively there may be another intermediate target of Dbt . Prior work in mammals has demonstrated a link between reduced circadian clock function and neurodegeneration , as well as a link between CKIδ/ɛ and apoptosis [42–45] . Reduced clock function has been produced by alterations to circadian transcriptional regulators , with increased neurodegeneration produced in response to reactive oxygen species and induction of apoptosis . CKI regulation of cell death and cell cycle arrest has been linked to effects on the mitotic spindle , p53 and cell surface receptors involved in cell death . It is likely that circadian clocks and CKI affect apoptosis and neurodegeneration at multiple steps in addition to the ones outlined in this manuscript . However , the direct effects of the clock components on Dbt levels and the consequent expression of an activated initiator caspase suggest that these events may be upstream and global mediators of circadian effects on apoptosis and neurodegeneration . It has been shown that these cell death components that are normally involved in destruction can also play critical roles in nonapoptotic events such as dendrite pruning , which occurs during development to create proper neural circuits . In Drosophila , caspase activity is detected locally in the degenerating dendrites and mutation of Dronc preserves most of the dendrite morphology [46] . In this instance caspases are not activated in the context of apoptosis , but in cell survival processes . A possible role for the Spag-Dbt-Dronc pathway in dendritic or axonal pruning/remodeling is intriguing in light of the existing literature on the role of the circadian clock and light in these processes[47–50] . There is circadian remodeling of lateral neuron ( PDF+ ) axon branching patterns as well as the size and synapses of several noncircadian neurons in the optic lobes , which are extensively innervated by the PDF+ axons . These circadian changes require a functional circadian clock , are enhanced by light , and in the case of some optic lobe changes require signals from the lateral neurons . Intriguingly accumulation of activated Dronc in the optic lobe at ZT7 in timGAL4>UAS-spag RNAi ( or UAS-dbtK/R ) flies also required light and was produced by PDF signaling from the lateral neurons; therefore , this activation may in fact be due to hyperactivation of the same pathways that trigger normal circadian neuronal remodeling in the optic lobes . In wild type flies , some of the optic lobe neurons exhibit largest axon size in the morning and the evening , suggesting that caspases ( presumably below the level of detection ) might contribute to pruning during the middle of the day , at times when activated caspases are detected here . It is not certain whether expression of activated Dronc in the optic lobe cells not expressing the spag RNAi ( or dbtK/R ) also involves reductions in Dbt activity in those cells or is produced by a different pathway in response to PDF signaling , but it is likely that Dbt reductions in these cells also occur , as the reductions in Dbt detected in immunoblots of total head extracts can be quite complete . We would argue that PDF signaling is important for global DBT reductions , casapse activation and Tau cleavage , and that these are produced cell autonomously in S2 cells by spag RNAi or DBTK/R expression . Reductions in spag may also trigger the non-cell autonomous reductions in Dbt levels and caspase activation , or these may be triggered by a spag-independent mechanism in response to spag decreases in the PDF+ cells . Caspase involvement in synapse degeneration has previously been suggested to contribute to AD pathologies [51] . We have identified a new mechanism of Tau cleavage and AD . We propose a model in which Spag regulates Dbt levels by regulating Dbt ubiquitination and/or phosphorylation , and when removed leads to the targeted degradation of Dbt . This removal of Dbt removes the inhibition on the caspase Dronc , leading to accumulation of its activated form and targeted cleavage of Tau ( Fig 7C ) . This is the first identification of a mechanism activating caspases in the context of AD and other tauopathies and sheds new light on the underlying mechanism that regulates the disease state and its connection to the circadian clock . Recently , we identified another TPR-containing protein ( Bride of Dbt , or Bdbt ) that interacts with Dbt to enhance Dbt activity and regulate its phosphorylation state [29] . It will be interesting to determine if this protein is part of any Spag-Dbt complexes that might regulate AD and apoptosis . Furthermore , it will be important to establish the genetic , environmental or aging processes that could interact with the mechanism to activate it during normal aging to produce AD or potentially other outcomes that negatively impact health and lifespan .
Drosophila RNAi stocks for spaghetti ( CG13570 ) were obtained from the Bloomington Drosophila Stock Center ( stock number: 31253 , targeting nucleotides 778–1200 of the transcript ) and the Vienna Drosophila RNAi Center ( VDRC stock numbers: 23896 , targeting nucleotides 886–1233 of the transcript; 103353KK , targeting nucleotides 780–1200 ) . In addition , the dronc RNAi line 23033 was obtained from the VDRC . The expression of hTau was under the control of the glass ( gl ) promoter provided by George Jackson [17] for Fig 5D and S1 Table , or alternatively a UAS-hTau fly line from Mel Feany [16] was used to express human Tau with the GMR-GAL4 driver on the X chromosome ( Bloomington stock center ) . The UAS-dbt-myc lines and the timGAL4 and UAS-dcr; timGAL4 driver lines have been described previously [27 , 28] . The pdfGAL4 driver line ( stock number 6899 ) and the Pdf receptor ( CG13758 ) null mutant line ( Pdfr5304; line 33068 ) were obtained from the Bloomington Drosophila Stock Center . ClkJrk [52] , pero , perS and perL mutations [53] were used alone or together with the dbtK/R constructs as described in the text . UAS-spag RNAi lines were crossed with UAS-dcr2; timGal4 lines . Progeny were then continuously reared at 23°C in a 12 hr:12 hr LD cycle for one more week after collection of adults ( or longer where indicated ) to insure complete RNAi knock-down effects , and loaded onto monitors ( Trikinetics , Waltham , MA ) for behavioral monitoring in constant darkness ( DD ) for at least 5 days . Actogram activity records and periodogram analysis to determine periods were employed as previously described with ClockLab [27] . Fly heads were collected at ZT 1 , 7 , 13 , and 19 or S2 cells at the times indicated in 1X SDS loading buffer , homogenized and stored at -80°C until analyzed . Samples were subjected to SDS-PAGE , transferred onto nitrocellulose and probed with appropriate antibodies: anti-DbtC ( 1:2000 ) [27] , anti-activated Dronc ( 1:100 ) [26] , anti-Actin ( 1:1000 ) ( Developmental Studies Hybridoma Bank ) , anti-Diap1 ( 1:100 ) [54] , anti-Tau ( Developmental Studies Hybridoma Bank ) and anti-HA ( 1:500 ) ( Covance PRB-101P ) , and signals were detected with the appropriate secondary antibodies and the ECL detection procedure . Anti-activated Dronc was purified by protein affinity purification . Antisera were applied to a protein G bead column bound with full length inactive Dronc and the flow-through was collected . This was followed by incubation of the flow-through with protein G beads bound with active Dronc . The beads were eluted with 0 . 1M glycine , pH 2 . 7 and used . Brains were dissected from flies at ZT7 and ZT19 , fixed , permeablized and incubated with anti-Dronc ( 1:100 ) , anti-CM1 ( Cell Signaling #9661 ) , anti-PDF ( 1:5000 ) ( PDF C7 , Developmental Studies Hybridoma Bank ) or anti-Myc ( 1:1000 ) ( Santa Cruz sc-789-G ) antibody overnight at 4°C . The following day samples were incubated with fluorescently-labeled secondary antibodies ( anti-rabbit IgG Alexa fluor 488 or anti-mouse IgG Alexa fluor 568 , 1:1000 ) . The brains were examined using an Olympus Fluoview confocal , and Z stacks were obtained . Drosophila S2 cells were transiently transfected with dTau-HA and 24h later cells were treated with dsRNA corresponding to dronc , dbt , spag or drIce and harvested 48h later . For UV treatment , cells were harvested 24h later . Samples were analyzed by immunoblot using anti-HA antibody . For treatment with active Dronc-His , S2 cells were harvested 24h after dTau-HA transfection , resuspended in Buffer A ( 25mM Tris·Cl pH 8 . 0 , 50mM NaCl , 10mM DTT ) and lysed by 4x freeze-thaw . After lysis , cells were spun down at 13000rpm and the supernatant was collected and incubated with active Dronc-His [54] for 4h at 30°C . The reaction was stopped by the addition of 5X SDS loading buffer and analyzed for dTau-HA by immunoblot analysis . As a positive control , extracts were analyzed from S2 cells expressing dTau-HA and treated with UV light . Fly eye size was measured from photomicrogaphs using the ImageJ program ( open source program , NIH ) . A circumference was drawn around the eye and the area was obtained from the measure command . The values ( number of pixels ) were then tabulated and averaged and statistics were performed using the Statistica ( Statsoft OK ) software . Fly heads ( 50–100 ) or S2 cells were collected and frozen in liquid nitrogen . Total RNA was isolated using the Trizol Reagent ( Invitrogen ) , and cDNA synthesis was performed using the Taqman Reverse Transcription Kit ( Life Technologies ) . PCR was performed using gene specific primers for spag and actin for 20 cycles , and the products were analyzed on an agarose gel . S2 cells stably expressing Spag or Dbt were untreated or treated with dsRNA corresponding to spag or dbt . Forty eight hours after dsRNA addition cells were harvested and analyzed for active Dronc , Dbt or actin . Plasmids used for this experiment were the following: pMT-dbt-myc , pMT-dbtK/R-myc , pAC-spag-ha . dbt ( nt 321–674 ) : forward 5’-TAATACGACTCACTATAGGGGCGCGTGGGTAACAAATATC-3’ , reverse 5’-TAATACGACTCACTATAGGGTGTATGTAATCGATGCGGGA-3’ , dronc ( nt 939–1454 ) : forward 5’-TAATACGACTCACTATAGGGATGGTGGGGATAGTGCCATA-3’ , reverse 5’-TAATACGACTCACTATAGGGTGTCAGGCCACTTCTCCTCT-3’ , drIce ( nt 653–968 ) : forward 5’- TAATACGACTCACTATAGGGACTGCCGCTACAAGGACATT-3’ , reverse 5’- TAATACGACTCACTATAGGGGCGTGCACTGGAATCTTGTA-3’ , diap1 ( nt 491–1019 ) : forward 5’-TAATACGACTCACTATAGGGCCGGCATGTACTTCACACAC-3’ , reverse 5’-TAATACGACTCACTATAGGGTTCTGTTTCAGGTTCCTCGG-3’ , spag set 1 ( nt 735–1242 ) : forward 5’-TAATACGACTCACTATAGGGCAAAAGTGGGCCAAACTTTAC-3’ reverse 5’TAATACGACTCACTATAGGGTTCTGGGCTGCGTTCTAT-3’ , spag set 2 ( nt22-227 ) : forward 5’-TAATACGACTCACTATAGGGGGAGCGCTAGCAACAGAAAT-3’ reverse 5’-TAATACGACTCACTATAGGGGCGACTTCTGGAGCTCTTTC-3’ , PCR fragments were produced from Drosophila genomic DNA with these primer sets . dsRNAi was produced and transfected into S2 cells by the procedures of the Perrimon lab ( http://www . flyrnai . org/DRSC-HOME . html ) , using the T7 promoters encoded by the primers . The cells were harvested at the times indicated in the figure legends . In one experiment , the proteasome inhibitor MG132 was added to a concentration of 50–100 μM , and in another MYC-tagged Dbt was overexpressed . A cDNA clone for CG13570 ( RE03224 ) was obtained from the Drosophila Genomics Resource Center ( DGRC , Bloomington , IN ) . We used the DGRC Gateway collection in order to clone spag into vectors allowing expression in S2 cells . A full-length spag ORF was generated by PCR ( Phusion , cat# F-553S , NEB ) and cloned into pENTR/SD/D-TOPO ( Cat# K240020 , Life Technologies ) by TOPO-mediated cloning . This clone was used to generate plasmids containing an HA tag . The clonase enzymatic mixture ( Cat# 11791–019 ) was purchased from Life Technologies . spag was cloned into the pGateway vector pAHW ( Cat#1095 ) to generate an N-terminal 3xHA-tag plasmid . The plasmid were driven by an Act5c promoter allowing for constitutive expression in S2 cells . For the expression of Drosophila Tau ( dTau ) in S2 cells , the pFLC1-TAU cDNA ( Drosophila Genomics Resource Center , Clone # RE16764 ) was cloned into the pAc S2 cell expression vector . The N-terminal HA tag was added using the following primers: Tau N-TER Forward , 5’-GCGCGAATTCGCGACCCTATGGCGTACCCGTACGACGTGCCGGACTACGCGATGGCGGATGTCCTGGAGAAAAGCTCACTG 3’ , Tau N-TER DRA Reverse , 5’—GGCGCATGTCCGACTTGTACC 3’ The DNA was digested with DraIII and EcoRI and swapped for the original fragment in the pAc-Tau . Flies were maintained at 24°C on a 12 hr:12hr light/dark cycle . For aging studies , virgin male flies were isolated and maintained in 20 per vial and transferred to a fresh vial every 10 days . The number of dead flies was recorded . For the climbing assay , flies were sorted into groups of ten per vial and tested . A climbing apparatus was prepared for each group , with two empty polystyrene vials vertically joined by tape facing each other . For the lower vial , we measured a vertical distance of 8 cm above the bottom surface and marked each vial by drawing a circle around the entire circumference of the vial . We transferred a group of ten flies into the lower vial and allowed the flies to acclimatize to the new setting for 1 minute before conducting the assay . Flies were gently tapped down to the bottom of the vial and the number of flies that climbed above the 8 cm mark by 10 seconds after the tap was measured[55] . This was repeated with the same group ten times , allowing for a 1 minute rest period between each trial . The average number of flies per group that passed the 8-cm mark as a percentage of total flies was recorded . | Alzheimer’s disease is the most common cause of dementia in the aging population . It is a progressive neurodegenerative disorder that attacks the brain neurons , resulting in loss of memory , thinking and behavioral changes . One pathological hallmark is aggregation of the microtubule-associated protein Tau . A growing body of evidence highlights the importance of caspase-dependent Tau truncation in initiation and potentiation of Tau aggregation . Here we use the fruit fly Drosophila to examine the links between circadian rhythms , aging , apoptosis and Alzheimer’s Disease . We identified a regulator ( spag ) of the circadian kinase Dbt that functions to stabilize Dbt during the middle of the day . In addition , the caspase Dronc is regulated by Dbt and Spag and , when activated by reduction of either , targets Tau for cleavage , leading to behavioral deficits and shortened lifespans . The expression of activated caspase occurs in several parts of the brain in a manner requiring signaling from a neuropeptide produced by circadian cells . Wild type flies with no genetic modifications eventually exhibit modified Dbt and expression of activated caspase at specific times of day , further demonstrating the links between the circadian clock , light and apoptosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Drosophila Spaghetti and Doubletime Link the Circadian Clock and Light to Caspases, Apoptosis and Tauopathy |
Pilocytic astrocytoma ( PA ) is the most common brain tumor in children . This tumor is usually benign and has a good prognosis . Total resection is the treatment of choice and will cure the majority of patients . However , often only partial resection is possible due to the location of the tumor . In that case , spontaneous regression , regrowth , or progression to a more aggressive form have been observed . The dependency between the residual tumor size and spontaneous regression is not understood yet . Therefore , the prognosis is largely unpredictable and there is controversy regarding the management of patients for whom complete resection cannot be achieved . Strategies span from pure observation ( wait and see ) to combinations of surgery , adjuvant chemotherapy , and radiotherapy . Here , we introduce a mathematical model to investigate the growth and progression behavior of PA . In particular , we propose a Markov chain model incorporating cell proliferation and death as well as mutations . Our model analysis shows that the tumor behavior after partial resection is essentially determined by a risk coefficient γ , which can be deduced from epidemiological data about PA . Our results quantitatively predict the regression probability of a partially resected benign PA given the residual tumor size and lead to the hypothesis that this dependency is linear , implying that removing any amount of tumor mass will improve prognosis . This finding stands in contrast to diffuse malignant glioma where an extent of resection threshold has been experimentally shown , below which no benefit for survival is expected . These results have important implications for future therapeutic studies in PA that should include residual tumor volume as a prognostic factor .
Pilocytic astrocytoma ( PA ) is the most common pediatric brain tumor and the second most frequent tumor in childhood [1] . Three of four cases are diagnosed up to an age of 20 years with the highest age incidence between 5 and 15 years . PA is usually benign , often follows an indolent course and is mostly slow-growing [2] . In children , PA most frequently occurs in the cerebellum but can develop in the entire neuroaxis . Surgery is the treatment of choice [3] . If total excision is achieved , the prognosis is favorable with more than 90% of patients being cured [4] . However , in many cases tumor location in critical or deep areas ( such as brain stem , optic pathway , or hypothalamus ) restricts resection options and alternative management options are required [5 , 6] . Patients with only partial resection have a worse and highly unpredictable prognosis [4 , 5] . Tumors can regrow or even progress to a more aggressive tumor [3 , 7–11] but spontaneous tumor regression of PA has also been observed [4 , 12–15] and is a common phenomenon . A recent review in [14] estimates a fraction of 14% of all residual cerebellar astrocytoma that regress spontaneously . Other studies claim an even higher portion [16] . While regression of PA after partial resection is reported in many case series [12–16] , the influence of the residual tumor size has not been evaluated yet . Moreover , the management for patients in whom complete resection cannot be achieved is still unclear . Due to the chance of regression and the indolent nature of PA , some authors propose a wait and see strategy in order to avoid potential risks induced by further therapies [4 , 7 , 14] . Other authors favor an aggressive surgical resection in combination with additional treatment strategies , like radiation and chemotherapy to control tumor growth [15 , 17 , 18] . On the molecular level , it has been shown that activation of the mitogen-activated protein kinase ( MAPK ) pathway is sufficient to induce the development of PA . This leads to the hypothesis that PA is a single-pathway disease [19 , 20] . Furthermore , PA usually harbor only one alteration within the MAPK pathway . The majority of mutations are activating changes in the BRAF gene , the most common is the KIAA1549-BRAF fusion , but also other activating mutations have been described . A more aggressive behavior of PA is observed if additional genetic alterations occur , e . g . loss of tumor suppressor gene CDKN2A [10 , 21] . Furthermore , alterations in the PI3K/AKT pathway [22] have been associated with aggressive forms of PA [9] . One proposed mechanism for the often observed slow growth of the tumors is oncogene-induced senescence , which is a mechanism limiting neoplastic growth by inducing cellular senescence . The MAPK activation might initially promote growth as well as induce senescence . Oncogene-induced senescence has also been observed in melanocytic nevi and melanoma [10] . Several mechanisms for tumor regression have been suggested , e . g . immunologic mechanisms , hormonal factors , induction of differentiation or apoptosis [13] . However , the reason why regression in PA occurs is not understood yet [4] . We formulate a mathematical model for growth , progression and regression of PA based on the above described clinical and molecular biological observations . We study the effects of competition between tumor and wild-type cells on the chance for regression . We distinguish two types of PA . Benign cases are classified as PA-I tumors and assumed to be caused by alteration of a single pathway . Tumors in which an additional alteration occurs are categorized as PA- II tumors , representing the more aggressive subset of PA . We introduce a stochastic tumor growth and progression model , namely a Moran model [23] with mutations . We chose a Moran model in this juvenile tumor , since astrocyte proliferation and diversification mainly happen during late embryogenesis and the first three weeks after birth . These processes are largely complete by early postnatal stages , while early and late postnatal development is mainly characterized by maturation processes ( like continuing elaboration of astrocyte processes and building of synaptic/vascular connections ) [24 , 25] . Since PA are usually diagnosed between 5 and 15 years , the normal astrocyte population is not proliferating at this time anymore . Therefore , it is reasonable to assume an approximately homeostatic tissue . In such a tissue , Moran dynamics provide a natural and established framework for modeling competition between tumor and wild-type cells . In our model , we derive the PA-regression-function describing the probability for regression in dependency of the residual tumor size after partial resection of benign PA . The accumulation of mutations in a tissue has been modeled and investigated by several authors by using a Moran model . Work by Iwasa , Michor , Komarova and Nowak [26 , 27] has been extended by Schweinsberg [28] and durrett , Schmidt and Schweinsberg [29] to the case of m mutations . These models analyze tumor growth and progression [30–34] with a focus on theoretical results regarding the waiting time until a cell has accumulated a certain number of mutations . Our approach is motivated by a concrete clinical question which is the regression probability of a benign PA tumor in dependency of the residual tumor size . We modify the model introduced in [29] . In particular , we consider Moran dynamics with two mutations but two absorbing states and investigate the precise relation of the two absorption probabilities which allows the incorporation of epidemiological data to calibrate the model . From the mathematical point of view , the relation of the two absorption probabilities can be connected to the portion of stochastic tunneling events in the model presented in [29] .
The behavior of the TGP process depends on its three parameters , the critical tumor size N , the mutation probability from wild-type cells to type-I cells u and the mutation probability from type-I cells to type-II cells v . The parameter regime for the analysis of the TGP model is chosen such that u ≪ 1 N ( 1 ) and ( N v ) 2 = : γ > 0 . ( 2 ) In the following we explain this choice . We call the parameter γ risk coefficient . We are interested in the regression probability of a partially resected PA-I tumor in dependency of the remaining tumor size and assume that regression of a residual tumor is achieved if no tumor cells are present anymore . All suggested mechanisms of tumor regression influence the ratio of tumor and wild-type cell birth and death rates . Therefore , we assume that competition between tumor and wild-type cells leads to tumor regression which is incorporated by Moran dynamics with relevant cell number equal to N again , see also Fig 3 . Furthermore , we assume that the partial resection reduces the residual number of PA-I cells below the critical tumor size N . Hence , the regression function is defined as the extinction probability of tumor cells , i . e . the probability to reach state 0 when starting the TGP process in some state k with 1 ≤ k ≤ N − 1 . For v = 0 , our TGP process simplifies to a neutral two-type Moran process in which the extinction probability is an established result and equals 1 - k N [32] . Here , we derive this extinction probability for our TGP process with three cell types . For the mathematical analysis , it is convenient to express this function in terms of ρ = k N . The fraction ρ describes the ratio between the residual number of PA-I cells after partial resection k , 1 ≤ k ≤ N − 1 , and the critical tumor size N . Formally , these considerations lead to the regression function β γ N ( ρ ) defined as β γ N ( ρ ) : = I P ( X t = 0 for some t ≥ 0 | X 0 = N ρ ) , ρ ∈ [ 0 , 1 ] . ( 4 ) Fig 3 provides a graphical representation of regression in the TGP model . A diffusion approximation of ( Xt ) t ≥ 0 leads to the Wright-Fisher diffusion process that can be utilized to approximate the term of Eq ( 4 ) . This approach was introduced in [29] and leads finally to a series representation as approximation of β γ N ( ρ ) . In S1 Fig it is shown that this series can be expressed by Bessel functions I n , n ∈ I N , [35] and that the regression function of the TGP model is given by β γ ( ρ ) = 1 - ρ I 1 ( 2 γ ( 1 - ρ ) ) I 1 ( 2 γ ) ( 5 ) for 0 ≤ ρ ≤ 1 . The graph of βγ is plotted in Fig 4C for different values of the risk coefficient γ .
The regression function Eq ( 5 ) depends on the parameters of the TGP model via the risk coefficient γ , see Eq ( 2 ) . This parameter is estimated such that the clinically observed fraction of PA-I tumors , denoted by p ^ , equals the theoretically obtained fraction α ( γ ) of absorption in state N in the TGP model . Subsequently , the derived risk coefficient is substituted into the regression function given by Eq ( 5 ) in order to obtain the specific PA-regression-function . Fig 4 summarizes the overall strategy of this approach . We estimate the clinically observed fraction of PA-I tumors on the basis of data reported in [10] . The authors analyzed 66 PAs with respect to their genetic profile and classified 57 cases as benign PA-I tumors and 9 cases as more aggressive PA-II tumors . This leads to p ^ = 57 66 = 0 . 8636 . In the TGP model , this clinically observed fraction corresponds to the absorption probability in state N , given by Eq ( 3 ) . Therefore , we set α ( γ ^ ) = p ^ = 0 . 8636 . This equation allows to calculate the risk coefficient γ ^ which yields γ ^ = 0 . 152 . Substituting γ ^ = 0 . 152 into the regression function given by Eq ( 5 ) allows to derive the PA-regression-function given by β 0 . 152 ( ρ ) = 2 . 3795 1 - ρ I 1 ( 0 . 7797 1 - ρ ) , 0 ≤ ρ ≤ 1 . ( 7 ) A plot of this function is provided in Fig 4C . This figure shows that the regression function is very robust to small alterations with respect to p ^ . Note that the actual risk coefficient may be smaller than the estimated value γ ^ = 0 . 152 due to the following considerations . The parameter N in our model represents a critical tumor size above which tumor regression cannot be expected anymore . However , the number of mutated cells in a diagnosed PA-I tumor may be larger than N because tumors could grow beyond this critical size without symptoms or due to a diagnostic gap between first symptoms and diagnosis . Therefore , a PA-I tumor can consist of more than N type-I cells and should have been more susceptible for progression to PA-II than accounted for in our TGP model . Hence , the risk of progression in our TGP model and therefore γ ^ might be overestimated . However , this would not change the linear dependency between residual tumor size and regression probability which is discussed in the following . We can show that the PA-regression-function Eq ( 7 ) is approximately linear by utilizing a Taylor expansion using Eq ( 6 ) . Substituting the estimated risk coefficient of the PA-regression-function γ ^ = 0 . 152 into Eq ( 6 ) leads to T 1 ( ρ ) = 0 . 9817 - ρ , ρ ∈ [ 0 , 1 ] . ( 8 ) This is a very good approximation since the remainder term can be estimated by | R 1 ( ρ ) | ≤ γ 8 = 0 . 152 8 = 0 . 0185 ( 9 ) for ρ ∈ [0 , 1] . Hence , the deviation of the PA-regression-function from the linear function T1 ( ρ ) is very small . Moreover , if the risk coefficient was overestimated , an even smaller deviation would be observed as Eq ( 9 ) implies . In order to provide a quantitative prediction of the regression probability given the absolute residual tumor size , we estimate the critical tumor size N in our model . Since the total cell number corresponds to the the tumor volume , we can interpret N also as minimum absolute tumor volume above which tumor regression cannot be expected anymore . The existence of this critical tumor size and its estimate of a cell number corresponding to a volume of 9 cm3 is justified in the following way . First , an extensive literature research indicated that tumor regression for residual cerebellar PA over 9 cm3 has not been reported yet , see S1 Table . Second , the prediction for patients with 78 cerebellar astrocytoma , including 62 PAs , has been investigated in [15] . Fig . 6 in [15] implies that the theoretical proportion of progression-free patients based on a Cox regression analysis with a residual tumor of 9 cm3 is estimated to be zero in the long-term . Finally , in [18] , the role of the extent of resection in the long-term outcome of low-grade gliomas is investigated including 93 PAs . It is stated that “‘the predicted outcome for patients is negatively influenced by even residual tumor volumes on the order of 10 cm3”’ . Incorporating the estimation for the critical tumor size of 9 cm3 into the PA-regression-function Eq ( 7 ) allows to quantify our predictions , indicating that any volume reduction of one cm3 below the critical size will add 10% to the chance for regression ( see also Fig 5 and Table 1 ) . In malignant brain tumors it has been shown that there is an EOR threshold below which no survival advantage is provided , e . g . in glioblastoma this threshold is 78% [37] . The existence of different tumor zones which basically reflect tumor heterogeneity is one proposed reason for such a threshold in malignant brain tumors [38] . In contrast , our results suggest the non-existence of such a threshold in PA . This is an immediate consequence of the linear dependency between residual tumor size and regression probability . If the residual tumor is smaller than the critical tumor size N , which marks the volume for which regression cannot be expected anymore , any reduction of the tumor volume will contribute to the regression probability . Importantly , this behavior stands in contrast to a non-linear dependency which would have been obtained in our model for a higher estimated risk coefficient γ , see Fig 5 .
In order to gain insights into the regression behavior after partial resection of benign PA , we introduced a stochastic TGP model based on recent molecular findings , functional , and clinical data . We derived a regression function that depends on the risk coefficient γ and quantifies the probability of regression in dependency of the residual tumor size . By incorporating epidemiological data on the clinically observed fractions of PA-I and PA-II cases , we estimated γ and derived the specific PA-regression-function , given by Eq ( 7 ) . The estimated PA-regression-function implies an approximately linear dependency between the residual critical tumor fraction and the regression probability as illustrated in Fig 4C . This linear dependency is supported by a Taylor approximation and an estimation of the remainder term , given by Eqs ( 8 ) and ( 9 ) , respectively . Furthermore , we quantitatively predicted the chance for tumor regression for benign PA by estimating the critical tumor size N , see Table 1 . Our TGP model incorporates assumptions based on clinical observations . It is observed in the clinics that PA-I tumors grow slowly , arrest in growth , or even regress . Hence , type-I cells in our model proliferate without fitness advantage . Furthermore , we assume that the first type-II cell that occurs leads to an aggressive form of PA , corresponding to malignant progression in PA . Alternatively , one could assign a success probability s to an emerging type-II cell , which represents the probability that a single type-II cell leads to a PA-II tumor . However , it has been shown in [36] that this is equivalent to considering an analog process with type-II mutation probability sv instead of v . This alternative process would lead to the same estimated risk coefficient γ ^ . Therefore , the estimated PA-regression-function would not change since this function is determined only by γ ^ . Further , we use asymptotic results for N → ∞ in order to calculate the theoretical portion of PA-I and PA-II cases in the TGP model . This is justified by the fact that a tumor consists of billions of cells . Simulation results given in S2 Table support these asymptotic results . They show that excellent accordance with formulas for finite N is reached even for small values of N . Moreover , we could show that the model is robust against small changes in the proportion of PA-I versus PA-II tumors as shown in Fig 4C . This robustness is an important property of the model since the proportion of PA-I can vary between different studies , especially since the sample size is often very small [39 , 40] . To our knowledge , the proposed TGP model is the first theoretical attempt to predict the regression behavior of PA . In particular , we analyzed PA regression based on the population dynamics of tumor and wild-type cells . The ratio of tumor cell birth and death rates is influenced by immunologic mechanisms , hormonal factors , induction of differentiation , or apoptosis , which could all contribute to tumor regression [13] . Since PA-I tumors grow very slowly , we assumed identical birth and death rates of type-I cells in our model . Our findings have clinically relevant implications . There is still controversy about the best treatment strategy for PA . Since PA is a slowly growing tumor and might even spontaneously regress , a wait and see strategy is an option besides more aggressive treatment strategies like radiation and chemotherapy . The decision for a more radical therapy would depend on the risk for recurrence ( or even progression ) and the chance of regression . However , long-term follow-up data about the probability of regression or progression after partial resection of PA is restricted and only retrospective studies with small case numbers are available [39–42] . The linear dependency between residual tumor size and regression probability in our model implies that every resected percentage point of a PA-I tumor contributes equally to the regression probability . Hence , there is no EOR threshold , but any small reduction in tumor mass provides an improvement in prognosis by increasing the probability for tumor regression . This prediction suggests a fundamentally different treatment strategy for PA compared to glioblastoma for which such a threshold has been determined [37] . Therefore , our results indicate that resection of a tumor should be aimed at even if a complete resection may not be possible . This is supported by studies showing that in patients with PA outcome depends on the extent of resection , although these studies only differentiate between biopsy , partial , subtotal , and gross/total resection and do not measure tumor volumes [3 , 15 , 16] . Moreover , if complete resection cannot be achieved , our results predict that the outcome linearly depends on the residual tumor volume . If there is a reasonable chance for regression of the residual tumor , it might be less justified to accept side effects by further therapies like radiation . This is an important result since the role of additional radiation therapy in treating children with tumors is highly controversial [8] . Unfortunately , as far as we know , there are no clinical studies on treatment of PA that take into account the influence of the residual tumor volume on patient outcome . We suggest that the residual tumor volume is an important prognostic marker and that a lack of sufficient volumetric data could be a reason for different results in clinical studies on additional treatment in PA . The results of this work should be further supported by future clinical studies that include volumetric data , which will improve the quantitative prediction of our model and form a statistical basis for clinical decision rules . | The most common brain tumor in children and young adults is pilocytic astrocytoma ( PA ) . This tumor is usually benign and often follows an indolent course . The treatment of choice is resection and the prognosis is very favorable if total excision can be achieved . However , due to the location of the tumor , only partial resection is possible in many cases . Partially resected PA could spontaneously regress , regrow or even progress to a more aggressive type of PA . We develop a mathematical model which describes the growth , progression and regression of PA . We are able to quantitatively predict the chance for regression in dependency of the remaining tumor size . This prediction has the potential to provide decision support to clinicians after partial resection of benign PA . Furthermore , our results imply that there is no resection threshold for PA below which no survival advantage is provided . This finding stands in contrast to malignant brain tumors where such a threshold has been experimentally shown . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Model-Based Evaluation of Spontaneous Tumor Regression in Pilocytic Astrocytoma |
In bacteria , transformation and restriction-modification ( R-M ) systems play potentially antagonistic roles . While the former , proposed as a form of sexuality , relies on internalized foreign DNA to create genetic diversity , the latter degrade foreign DNA to protect from bacteriophage attack . The human pathogen Streptococcus pneumoniae is transformable and possesses either of two R-M systems , DpnI and DpnII , which respectively restrict methylated or unmethylated double-stranded ( ds ) DNA . S . pneumoniae DpnII strains possess DpnM , which methylates dsDNA to protect it from DpnII restriction , and a second methylase , DpnA , which is induced during competence for genetic transformation and is unusual in that it methylates single-stranded ( ss ) DNA . DpnA was tentatively ascribed the role of protecting internalized plasmids from DpnII restriction , but this seems unlikely in light of recent results establishing that pneumococcal transformation was not evolved to favor plasmid exchange . Here we validate an alternative hypothesis , showing that DpnA plays a crucial role in the protection of internalized foreign DNA , enabling exchange of pathogenicity islands and more generally of variable regions between pneumococcal isolates . We show that transformation of a 21 . 7 kb heterologous region is reduced by more than 4 logs in dpnA mutant cells and provide evidence that the specific induction of dpnA during competence is critical for full protection . We suggest that the integration of a restrictase/ssDNA-methylase couplet into the competence regulon maintains protection from bacteriophage attack whilst simultaneously enabling exchange of pathogenicicy islands . This protective role of DpnA is likely to be of particular importance for pneumococcal virulence by allowing free variation of capsule serotype in DpnII strains via integration of DpnI capsule loci , contributing to the documented escape of pneumococci from capsule-based vaccines . Generally , this finding is the first evidence for a mechanism that actively promotes genetic diversity of S . pneumoniae through programmed protection and incorporation of foreign DNA .
While sexual reproduction , which is crucial for genetic diversity in eukaryotes , is lacking in bacteria , genetic transformation is regarded as a substitute [1] . Genetic transformation proceeds through the internalization of single stranded ( ss ) DNA fragments created from an exogenous double stranded ( ds ) DNA substrate , which are incorporated into the genome by homology . This forms a heteroduplex of one strand of host DNA associated with complementary exogenous ssDNA , which is resolved by replication , producing one wild-type daughter chromosome , and one possessing the mutation originally present on the exogenous DNA . This widespread process [2] contributes to genetic plasticity of the major human pathogen Streptococcus pneumoniae ( the pneumococcus ) [3] , potentially leading to antibiotic resistance acquisition and vaccine escape [4] . Current pneumococcal vaccines target the polysaccharide capsule , which is considered the main pneumococcal virulence factor , and of which over 90 different serotypes exist [5] . Cumulatively , these serotype loci are almost equivalent in size to a single pneumococcal genome , demonstrating the high levels of genetic diversity present in the pneumococcal population . Vaccine escape presumably occurs via exchange of capsule loci , ranging in size from 10 . 3 kb to 30 . 3 kb , by transformation . These loci occupy the same position in the genome , between the dexB and aliA genes which provide flanking homology for chromosomal integration of heterologous capsule sequences . On the other hand , many bacteria possess DNA restriction-modification ( R-M ) systems , which defend against invasion by foreign DNA such as bacteriophage , and are seen as restrictive to genetic diversity , potentially antagonizing genetic exchange processes . R-M systems classically encode a restrictase , which degrades unmethylated ( me0 ) foreign dsDNA , and a methylase which methylates the dsDNA of the host genome , protecting it from the restrictase [6] . In the course of transformation and despite the widely accepted notion that R-M systems antagonize genetic exchange processes , internalized me0 ssDNA is resistant to restriction as most restrictases cannot cut ssDNA . Upon transfer of a point mutation , heteroduplex formation produces hemimethylated ( me+/0 ) dsDNA , which remains resistant to restriction , but can be methylated by the dsDNA methylase ( Fig . 1A ) . Resolution of the heteroduplex by replication then forms two me+/0 dsDNA , which again remain resistant to restriction . As a result , at no point during the transfer of a point mutation is the internalized or chromosomally-integrated DNA sensitive to restriction . The situation is potentially different for the transfer of heterologous DNA , since the heterologous region remains in the form of a ss loop after heteroduplex formation , flanked by regions of homology ( Fig . 1BC ) . In the case of methylated ( me+ ) transforming DNA , this should not be problematic , as the integrated ssDNA is already methylated , confering resistance to restriction to the dsDNA formed at the region of heterology after resolution of the heteroduplex by replication ( Fig . 1B ) . However , in the case of me0 transforming DNA , resolution of the heteroduplex by replication creates me0 dsDNA , which should be sensitive to restriction ( Fig . 1C ) . As a result , classical R-M systems should severely limit the acquisition of heterologous sequences of me0 origin . S . pneumoniae strains possess one of two complementary R-M systems , DpnI or DpnII [7] , encoded at a common location in the chromosome [8] . DpnI is an atypical restriction enzyme , as it cleaves me+ dsDNA . Chromosomal DNA produced by replication in DpnI cells is me0 and therefore is not sensitive to DpnI restriction . In contrast to this , DpnII strains possess a classical companion dsDNA methylase ( DpnM ) , which methylates host DNA after replication , rendering the chromosome resistant to the DpnII restrictase ( encoded by dpnB ) , while DpnII protects the cell from me0 bacteriophage attack by restricting me0 dsDNA . The dpnII locus also encodes DpnA ( Fig . 2 ) , a methylase with the unusual ability to specifically modify ssDNA ( ss-methylase ) . Whilst the locus is under the control of the constitutive Pdpn promoter , dpnA is also specifically induced from PX , a competence-inducible ( cin ) promoter recognized by the competence-specific σ factor , ComX [9] , [10] . This alternative σ is transiently active when pneumococcal cells become competent for genetic transformation [11] and required for synthesis and assembly of a dedicated complex for uptake and integration of exogenous DNA into the chromosome , known as the transformasome [12] . The primary role of DpnA was suggested to be protection of plasmids taken up by genetic transformation from DpnII restriction , via methylation of internalized plasmid ssDNA fragments [13] , [11] . Authors demonstrated that installation of me0 plasmids in S . pneumoniae was drastically reduced in a DpnII strain lacking DpnA , and concluded that DpnA biological role was to protect me0 plasmids from DpnII [13] . However , this appeared unlikely to be the function of DpnA , since plasmids naturally carried by S . pneumoniae are extremely rare [14] , [15] and plasmid transformation is strikingly inefficient in this species [16] . An alternative hypothesis was therefore proposed , that the role of DpnA was to protect heterologous me0 ssDNA to allow exchange of pathogenicity islands and , more generally , plasticity islands , defined as chromosomal regions variable between pneumococcal isolates [3] . According to this hypothesis , DpnA should methylate internalized transforming me0 ssDNA . Resolution of the transformation heteroduplex by replication should then result in formation of me+/0 rather than me0 dsDNA , thus protecting the transformed chromosome from restriction by DpnII . Thus , methylation of me0 transforming ssDNA by DpnA should be crucial for exchange of plasticity islands . Conversely , the absence of DpnA should leave the resulting transformants sensitive to DpnII ( Fig . 1C ) , which should restrict the chromosome , killing the cell . Recently , evidence was obtained that the inefficiency of plasmid transformation is presumably intrinsic to the transformasome , due to the competence-induced ssDNA-binding protein , SsbB . We showed that SsbB protects internalized ssDNA and creates a reservoir favoring chromosomal transformation [17] . In contrast , SsbB was found to antagonize plasmid transformation [17] , strongly suggesting that the pneumococcal transformasome has evolved to optimize chromosomal but not plasmid transformation . This conclusion prompted us to explore and validate the above-described hypothesis that DpnA is crucial for exchange of plasticity islands [3] . In this study , we show that in the absence of DpnA , acquisition of large me0 heterologous regions is drastically reduced . We also demonstrate that σX-dependent induction of dpnA during competence is required for full protection of me0 foreign DNA . We conclude that the competence-induction of dpnA and the methylation of me0 ssDNA by DpnA are crucial for protection of foreign me0 DNA , allowing exchange of pathogenicicy islands such as the capsule locus . We conclude that the recruitment of dpnA and dpnB ( encoding DpnII ) into the competence regulon provides the pneumococcus with protection from me0 bacteriophage through the action of DpnII , whilst maintaining the potential of acquisition of me0 pathogenicity islands ( e . g . , of DpnI origin ) via the protective role of ssDNA methylation by DpnA .
In order to validate our hypothesis that DpnA should be important for transfer of me0 heterologous pathogenicicy islands in DpnII strains , we created a series of isogenic strains possessing the full dpnII locus , or the dpnII locus with a previously characterized internal deletion of dpnA [13] ( Fig . 2 ) . We compared the transformation efficiency of cassettes on donor chromosomal DNA from either DpnI ( me0 ) or DpnII ( me+ ) origin in dpnA+ and dpnA− competent cells . We began by confirming a previous conclusion , based on indirect comparisons between non-isogenic pairs of donor and recipient cells , that DpnA plays no role in transformation of a chromosomal point mutation [13] using rpsL41 ( conferring streptomycin resistance , SmR ) [18] ( Fig . 3A ) . We then tested the efficiency of transformation of a cps2E::spc cassette from either DpnI ( me0 ) or DpnII ( me+ ) donor DNA into DpnII recipients with or without dpnA . This transformation required integration of ∼8 . 7 kb of heterology within the capsule locus . Efficiency was compared to that obtained with the rpsL41 SmR point mutation on the same donor DNA . As expected , there was no deficit in transformation efficiency in a dpnA− recipient when the donor DNA was me+ ( Fig . 3B ) . In contrast , a >4-log deficit in transformation efficiency was observed when the donor DNA was me0 ( Fig . 3C ) . This is because in presence of DpnA , prior methylation of ssDNA protects resulting transformants ( as in Fig . 1B ) , and in absence of DpnA , the incorporated DNA remains me0 , and resulting transformants are sensitive to DpnII restriction ( as in Fig . 1C ) . This result confirms that protection of heterologous me0 donor DNA from restriction by DpnA-mediated methylation is crucial for successful transfer . The cps2E::spc cassette represents a large heterology , with 19 GATC sites available for DpnII restriction . To determine whether DpnA was required to protect shorter regions of heterology , with fewer GATC sites , transfer of five further cassettes ( Table 1 ) was compared . The results , presented as ratio of transformation frequency in dpnA− cells compared to dpnA+ cells ( after normalization to that of the SmR point mutation present on the same donor DNA ) , termed dpnA−/+ ratio , show that longer heterologies are correspondingly more dependent on DpnA for protection ( Fig . 4A ) . Indeed , most of the transfers result in a deficit in dpnA−/+ ratio of 3 to 4 logs between me+ and me0 donor DNA , reinforcing the importance of DpnA for protection of these me0 heterologous cassettes . This relationship could simply reflect the increased probability of finding GATC sites with increasing heterologous segment length ( Fig . 4B ) . To establish whether the number of GATC sites is important in determining the dependence on methylation of ssDNA by DpnA , we modified the glnR::kan22C cassette to create derivatives with the same heterology length ( 1 , 336 bp ) but with 3 ( glnR::kan22C ( 3 ) ) , 6 ( glnR::kan22C ( 6 ) ) or 8 ( glnR::kan22C ( 8 ) ) me0 GATC sites . If the number of GATC sites present in the heterologous DNA determined the reliance of successful transfer on DpnA-mediated methylation of ssDNA , we expected that glnR::kan22C ( 3 ) should transfer into a dpnA− strain with higher efficiency than glnR::kan22C ( 6 ) , which should in turn transfer with greater efficiency than the wild-type glnR::kan22C ( 8 ) cassette . The results confirm this prediction , showing that the dpnA−/+ ratio of transformation frequencies is inversely proportional to the number of GATC sites in the heterology , with transformation efficiency in dpnA− recipient cells decreasing as GATC sites in donor heterology region increase ( Fig . 4C ) . The number of GATC sites present in the heterologous donor DNA thus determines the importance of DpnA , suggesting that the ss-methylase is particularly important for the transfer of long heterologous regions of me0 DNA . We propose that absence of DpnA results in direct competition between DpnM and DpnII for access to integrated me0 dsDNA following replication , where DpnM must methylate all GATC sites to protect the chromosome before DpnII restricts a single one . As a result , increasing the number of heterologous GATC sites favors DpnII in this competition , resulting in greater loss of transformant cells , since a greater proportion are destroyed by DpnII . Although it was previously shown that DpnA is specifically induced during competence for genetic transformation [11] , the specific induction profile of DpnA was not established . Prior to determining the importance of DpnA competence induction for protection of me0 heterologous transforming DNA , we further explored the expression of DpnA during competence induction . Firstly , ectopic luciferase reporter fusion assays confirmed that the wild type ComX-dependent promoter , PX , was induced by competence-stimulating peptide ( CSP ) [19] , whilst a promoter with a mutated cin box , PX- , was completely inactive ( Fig . 5A ) . Secondly , a time course Western-blot showed that DpnA was expressed in two forms , a longer form , weakly constitutively expressed from Pdpn ( DpnAL ) , and a shorter form specifically expressed from PX during competence ( DpnAS ) ( Fig . 5B ) . The observation of both DpnAL and DpnAS proteins in competent pneumococcal cells establishes that the two forms previously observed in E . coli extracts [13] are biologically relevant . These results demonstrate that although a small proportion of DpnA is constitutively produced , the vast majority of DpnA is produced during competence for genetic transformation . The importance of DpnA competence induction for pathogenicicy island exchange was investigated by mutating PX in the dpnII locus , and comparing the transfer efficiency of me0 cassettes in this recipient ( cin− dpnA+ ) to that in a dpnA+ recipient . Results showed that induction of DpnA during competence was crucial for full protection and transfer of me0 heterologous DNA , as transformation efficiency of me0 cassettes decreased substantially in the cin− dpnA+ recipient strain ( Fig . 5C ) . Furthermore , comparing the ratio of transfer between cin+ dpnA+ and cin− dpnA+ recipient strains showed that an increase in the number of heterologous GATC sites reinforced the importance of competence induction of DpnA ( Fig . 5D ) . As noted above , this is likely due to the fact that DpnM and DpnII compete directly for access to me0 GATC sites in the chromosome after integration of me0 heterologous DNA and replication . It is of note that mutating the cin box in an otherwise dpnA− recipient ( i . e . , rendering the cell cin− dpnA− ) increased transformation efficiency compared to cin+ dpnA− cells ( Fig . 5E ) . We attribute this increase to the non-induction of dpnB at competence , which results in lower levels of DpnII , shifting the competition between DpnM and DpnII for access to GATC sites in favor of the former , and thus increasing the probability that me0 dsDNA GATC sites will be protected after replication .
The primary role of the DpnII R-M system is to protect the pneumococcus from attack by me0 bacteriophage [13] . Here we show that the ss-methylase DpnA permits acquisition of me0 pathogenicicy islands ( e . g . , from DpnI pneumococci ) via methylation of internalized foreign heterologous me0 ssDNA ( Fig . 3C ) . Transformation of a 21 . 7 kb heterologous region is thus reduced by more than 4 logs in dpnA mutant cells . DpnA-dependent methylation occurring before as well as possibly after integration ( i . e . , at the heteroduplex stage; Fig . 1C ) renders the incorporated dsDNA resistant to DpnII after replication ( Fig . 1B ) , and ensures the survival of the transformed cell , allowing exchange of pathogenicicy islands between DpnI and DpnII populations . A time course Western-blot following CSP addition revealed that the vast majority of DpnA is produced during competence for genetic transformation ( Fig . 5B ) . The importance of competence induction of dpnA for acquisition of heterologous DNA was evaluated through inactivation of the cin box , bound by σX , revealing a ∼100-fold reduction in transformation for a segment harboring 19 GATC sites ( Fig . 5C and 5D ) . This reduction establishes that for such a large heterology ∼99% of the protection relied on DpnA molecules synthesized at competence . It is of note that when comparing heterologous transfers in dpnA+ and cin− dpnA+ recipients to assess the impact of dpnA competence induction , we likely underestimate the true importance of this induction . We observed that the transformation efficiency in a cin− dpnA+ strain remains higher than in a cin+ dpnA− equivalent ( compare Fig . 5C and 5E ) , and attribute this to two factors . Firstly , DpnAL remains ( weakly ) constitutively expressed in cin− dpnA+ strains . Secondly , we have shown that dpnB ( encoding DpnII ) , as well as dpnA , is part of the competence regulon ( Fig . 5E ) . As a result , mutating PX reduces not only DpnA but also DpnII levels , which should favor DpnM in the race against DpnII , allowing survival of more recombinants and resulting in underestimation of the effect of non-induction of DpnA during competence . Since the vast majority of DpnA is produced during competence ( Fig . 5B ) , we conclude that the induction of DpnA is critical for full protection of heterologous ssDNA internalized during competence . We have shown DpnA-mediated protection to be especially important in the case of large pathogenicicy islands ( Fig . 4A ) . The number of GATC sites , which are targeted by DpnII , being proportional to length of heterologous regions and the abundance of these sites dictating the probability of restriction by DpnII ( Fig . 4 ) readily explain this observation . The position of the PX promoter downstream of dpnM ( Fig . 2 ) results in induction of both dpnA and dpnB , but not dpnM , during competence . What could the biological relevance of this organization be ? A first advantage we see in the recruitment of dpnA to protect transforming DNA is that the window of opportunity for DpnA methylation of heterologous ssDNA is long , existing between internalization of exogenous ssDNA , and the passage of the replication fork over the recombinant chromosome . In contrast , DpnM can only methylate heterologous DNA in the form of dsDNA incorporated into the chromosome , a much smaller window of opportunity . Our findings confirm this interpretation by showing that DpnM-mediated protection is inefficient , since in the absence of DpnA , the vast majority of transformants are lost due to DpnII restriction ( Fig . 3C ) . Moreover , competence induction of dpnM ( i . e . , to favor protection versus restriction ) would be counter-productive for the cell , potentially increasing protection of me0 dsDNA phage and thus antagonizing the main role of DpnII . DpnA affords an elegant solution to this problem , as it provides protection to transformation intermediates via ss-methylation , but cannot antagonize DpnII-mediated phage destruction . Protecting long heterologous regions is of particular importance for the transfer of pathogenicity islands , such as the capsule locus which is generally considered the most important virulence factor of the pneumococcus . Capsule switching has been shown to be a prominent means of vaccine escape [20] . DpnA is likely to be crucial for DpnII cells to acquire serotypes from DpnI origin by transformation and , more generally , to allow evolution of the bacteria by acquisition and integration of foreign DNA . As concerns the specific co-induction of dpnB with dpnA during competence , if DpnII concentration increase in competent cells is similar to that observed for DpnA , then clearly it pushes the DpnII/DpnM ratio toward restriction thus increasing the protection against dsDNA me0 bacteriophage , whilst maintaining the plasticity potential of the cell via DpnA . The DpnI and DpnII R-M systems appear equally distributed in the pneumococcal population ( Table 2 ) . In the case of DpnI strains , no genetic barrier exists , as even when integrating me+ heterologous DNA ( i . e . , from DpnII ) , me+ dsDNA GATC sites are never produced in the host chromosome , which is never sensitive to DpnI restriction . In the case of DpnII strains , we have shown that DpnA protects heterologous me0 ssDNA , allowing acquisition of foreign pathogenicicy islands and preventing formation of a genetic barrier between the two populations . We would therefore predict that DpnI/DpnII R-M systems are not a barrier to genetic exchanges amongst pneumococci , i . e . , that other variable loci such as capsular serotype [5] , capsule regulatory module [21] or CSP pherotype [22] would be randomly distributed in the DpnI and DpnII populations . This prediction proved to be correct as evidenced by the 19A and 19F capsule serotypes in both Dpn types ( Table 2 ) . As a Dpn switch should be lethal , since expression of DpnI in a DpnII recipient would destroy the chromosome , and vice versa , we infer that the observed switches occurred in the other variable loci . The DpnI/DpnII split thus protects the pneumococcal species from both me+ and me0 bacteriophage , ensuring the survival of the pneumococcal species in the face of any bacteriophage attack , without creating an evolutionary barrier between the two sub-populations , thanks to the existence of and competence-induced production of DpnA . Although methylases specific for ssDNA such as DpnA appear to be rare , a few others exist . Both DpnI and DpnII are present in Streptococcus mitis ( our observations ) , a species closely-related to S . pneumoniae , and we propose that DpnB/DpnA play a similar dual role of protection from bacteriophage attack and maintenance of genetic exchange . Supporting this notion , the cin box in the PX dpnA promoter is perfectly conserved in a S . mitis DpnII strain [23] ( our observations ) , suggesting that dpnA and dpnB form part of the competence regulon in this species as well . In Bacillus centrosporus , an analogous R-M system , BcnI , also possessing an ss-methylase , was identified [24] . If B . centrosporus were transformable , it would be interesting to investigate whether this ss-methylase plays a similar role to that of DpnA with respect to pathogenicicy/plasticity island exchange . Competence for genetic transformation is widespread in bacteria [2] , although few studies have addressed the effect of R-M systems on genetic plasticity in these species . In the naturally transformable species Helicobacter pylori , R-M systems antagonize transformation , restricting genetic plasticity [25] . However , due to the internalization of exogenous DNA in ss form , the authors remained unclear as to the mechanism involved in this antagonization , and stated in a recent review ‘Because restriction enzymes prefer dsDNA , they likely act prior to unwinding and perhaps extracellularly’ [26] . Our model of DpnII-mediated restriction of transformant chromosomes formed after integration of me0 heterologous DNA ( Fig . 1C ) addresses this problem , highlighting the availability of fully sensitive me0 dsDNA in the transformant chromosome after replication , explaining how classical R-M systems can antagonize heterologous transformation . In contrast , we show that through programmed protection of foreign , heterologous me0 ssDNA by DpnA , the pneumococcal DpnII R-M system actively promotes genetic plasticity by favoring acquisition of heterologous me0 cassettes . This , along with the lack of identified ss-methylases in other transformable species , suggests that not all competent species have developed R-M systems that protect from bacteriophage attack , whilst at the same time maintaining the potential for genetic plasticity , and thus adaptation to environmental stresses by acquisition of virulence loci . It is likely that the need for such genetic flexibility is species-specific , since the pneumococcus , a pathogen and common commensal of humans constantly exposed to host , antibiotic and vaccine stresses , may need to maintain its propensity for rapid adaptation more so than another species occupying a different niche . The relative loss of genetic plasticity afforded by classic R-M systems may not be problematic for bacterial species that are not readily exposed to such selective pressures , and as such do not need to maintain a high level of genetic plasticity . Other systems of defense against attack by foreign DNA exist , such as the clustered , regularly interspaced short palindromic repeat loci ( CRISPR ) , which are sequence-specific defense mechanisms against bacteriophage , and have been shown to constitute a barrier to genetic diversity by horizontal gene transfer [27] . The pneumococcus does not possess CRISPR sequences , but a recent study artificially inserted CRISPR loci into the pneumococcal genome , and demonstrated that CRISPR interference blocked genetic plasticity by transformation in an in vivo model [28] . The authors concluded that CRISPR interference can limit the adaptability of a pathogen to a particular stress by limiting genetic plasticity . In contrast to our findings on the DpnII R-M system , no known mechanism exists to negate the antagonistic effect of CRISPR on genetic plasticity . Since we suggest that the pneumococcus appears to preferentially develop mechanisms that promote genetic diversity [3] , the absence of CRISPR loci in this pathogen is not surprising . Indeed , despite the negative effect of CRISPR on genetic plasticity , the pneumococcus demonstrated the genetic flexibility required to spontaneously eject the CRISPR sequences when under environmental stress in the in vivo model , thus permitting a number of clones to adapt and survive [28] . It is unclear whether this ejection would be possible in bacteria that naturally possess CRISPR sequences , and do not favor adaptation by genetic plasticity to the same extent as the pneumococcus . Bacterial pathogens must overcome two types of threat to survive and thrive , attack by foreign DNA ( in the form of bacteriophage ) , and defenses mounted by host factors , antibiotics and vaccines . The pneumococcal DpnII system is an elegant system achieving both these ends . The DpnII restrictase protects the cell from me0 dsDNA bacteriophage attack , whilst the DpnA ss-methylase maintains the pneumococcus' extraordinary potential for genetic plasticity , allowing adaptation to and evasion of defenses , natural or artificial , mounted by the human host upon pneumococcal infection . The question of why bacteria internalize exogenous DNA has been discussed at length [29]–[32] . The two credible suggestions are the use of DNA for genome maintenance via template-directed repair , or to promote genetic diversity via integration of exogenous DNA into the host genome . Here we show that competence induction of a locus coupling an ss-methylase to a restrictase is crucial for protection of heterologous exogenous DNA in the pneumococcus . We do not see how this protection could participate in genome maintenance , whereas it can clearly facilitate exchange of pathogenicity/plasticity islands . As a result , we consider this finding as the first direct evidence that S . pneumoniae can take up foreign DNA with the specific goal of integrating it into its chromosome via recombination , to promote genetic diversity .
S . pneumoniae strain growth and transformation were carried out as described [33] . Strain , plasmid and primer information can be found in Table S2 . Recipient strains were rendered hex− by insertion of the hexA::ermAM cassette as described [34] , negating any effect of the mismatch repair system on transformation efficiencies [35] . Antibiotics were used at the following concentrations; Chloramphenicol 4 . 5 µg mL−1 , Erythromycin 2 µg mL−1 , Kanamycin 250 µg mL−1 , Spectinomycin 200 µg mL−1 , Streptomycin 200 µg mL−1 . The dpnI locus was replaced with the dpnII locus in laboratory strain R800 by Janus [36] . Briefly , the dpnI locus from R800 was amplified by PCR with primers dpnCup and dpnDdo , possessing BamHI and PstI restriction sites respectively , and cloned into pGBDU . The resulting plasmid ( pGBDU-dpnI ) was digested by ClaI and SacI restriction enzymes , deleting a 180 bp internal fragment of dpnC , encoding the DpnI restrictase . The Janus cassette was amplified with primers kan5c and 7 sac , possessing ClaI and SacI sites , and ligated into pGBDU-dpnC− , to give pGBDU-dpnI-Janus . The resulting plasmid was transformed into R981 , with kanamycin selection , creating R2888 . The Janus cassette was replaced with dpnII locus either wildtype ( G54 donor DNA ) or dpnA− ( 1135 donor DNA [13] ) creating R2980 and R2981 , respectively . The resulting strains were rendered SmS by transformation with a wild-type rpsL PCR fragment to remove the SmR point mutation rpsL1 , creating R2992 and R2993 , respectively . To calculate dpnA−/+ ratios , the transformation efficiency of a specific cassette into dpnA− and dpnA+ recipients was divided by the number of transformants by total number of colony forming units ( cfu ) obtained . This value was then normalized against a point mutation transformed on the same donor DNA ( rpsL41 , conferring SmR ) , unaffected by absence of DpnA . The resulting transformation efficiency for a dpnA− recipient was divided by the same value for the same cassette in a dpnA+ recipient to give the dpnA−/+ ratio . A ratio of 1 indicates no effect of absence of DpnA on transformation efficiency , whilst a ratio of <1 indicates a loss of transformation efficiency in a dpnA− recipient strain . Further methods used in this study can be found in the Supporting Information . This file contains detailed descriptions of the cassettes used in this study , reporter fusions used to measure transcription of the PX promoter , creation of the cin− mutated dpnA promoter and construction of dpnA-SPA . | Natural genetic transformation can compensate for the absence of sexual reproduction in bacteria , allowing genetic diversification by recombination . It proceeds through the internalization of single stranded ( ss ) DNA fragments created from an exogenous double stranded ( ds ) DNA substrate , which are incorporated into the genome by homology . On the other hand , restriction-modification ( R-M ) systems , which protect bacteria from bacteriophage attack by degrading invading foreign DNA , potentially antagonize transformation . About half of the strains of the naturally transformable species and human pathogen Streptococcus pneumoniae possess an R-M system , DpnII , restricting unmethylated dsDNA . DpnII strains possess DpnA which is unusual in that it methylates ssDNA . Here we show that DpnA plays a crucial role in the protection of internalized heterologous transforming ssDNA , preventing the post-replicative destruction by DpnII of transformants produced by chromosomal integration of heterogolous DNA by virtue of flanking homology . This protective role of DpnA is of particular importance for acquisition of pathogenicity islands , such as capsule loci , from non-DpnII origin by DpnII strains , likely contributing to pneumococcal virulence via escape from capsule-based vaccines . Generally , this finding is the first evidence for a mechanism that actively promotes genetic diversity of S . pneumoniae through active protection and incorporation of foreign DNA . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2013 | Programmed Protection of Foreign DNA from Restriction Allows Pathogenicity Island Exchange during Pneumococcal Transformation |
Here we investigated the role of the Nod/Rip2 pathway in host responses to Chlamydophila pneumoniae–induced pneumonia in mice . Rip2−/− mice infected with C . pneumoniae exhibited impaired iNOS expression and NO production , and delayed neutrophil recruitment to the lungs . Levels of IL-6 and IFN-γ levels as well as KC and MIP-2 levels in bronchoalveolar lavage fluid ( BALF ) were significantly decreased in Rip2−/− mice compared to wild-type ( WT ) mice at day 3 . Rip2−/− mice showed significant delay in bacterial clearance from the lungs and developed more severe and chronic lung inflammation that continued even on day 35 and led to increased mortality , whereas WT mice cleared the bacterial load , recovered from acute pneumonia , and survived . Both Nod1−/− and Nod2−/− mice also showed delayed bacterial clearance , suggesting that C . pneumoniae is recognized by both of these intracellular receptors . Bone marrow chimera experiments demonstrated that Rip2 in BM-derived cells rather than non-hematopoietic stromal cells played a key role in host responses in the lungs and clearance of C . pneumoniae . Furthermore , adoptive transfer of WT macrophages intratracheally was able to rescue the bacterial clearance defect in Rip2−/− mice . These results demonstrate that in addition to the TLR/MyD88 pathway , the Nod/Rip2 signaling pathway also plays a significant role in intracellular recognition , innate immune host responses , and ultimately has a decisive impact on clearance of C . pneumoniae from the lungs and survival of the infectious challenge .
Chlamydophila pneumoniae is a Gram-negative obligate intracellular pathogen that is widely prevalent [1] , causes respiratory tract diseases such as pneumonia , sinusitis , and bronchitis , contributes to acceleration of atherosclerosis [2] , [3] , and is associated with development of chronic lung diseases such as asthma [4] and other disorders where chronic inflammation is a hallmark feature [5] , [6] . C . pneumoniae infects various cell types such as epithelial cells , monocytes , macrophages , smooth-muscle cells and endothelial cells , and often resides intracellularly for indefinite periods [7] . C . pneumoniae induces a similar lung pathology in humans and rodents [8] . A mouse model of lung infection has been used to study the immunological mechanisms of host defenses . Host immune responses to C . pneumoniae proceeds in two stages; 1 ) an early response requiring IFN-γ to limit the growth of the intracellular bacteria , which plays a central role in the innate control of this infection , and 2 ) a later adaptive immune response that includes CD4+ and CD8+ T cells in bacterial clearance and protection [9]–[11] . While the primary immune response is aimed to clear the primary infection from the host and provide protection against reinfection with the same pathogen , generation of tissue injury also occurs and Chlamydial infections often recur or remain persistent and long-term consequences of recurrent or persistent chlamydial infections can be severe [10] , [12] . Chlamydia is internalized by macrophages as well as by “non-professional” phagocytes , where it survives and replicates . C . pneumoniae elicits IFN-γ production in infected bone marrow-derived macrophages [13] . In such cells , IFN-γ synergizes with bacterial products to activate various bactericidal mechanisms , including inducible nitric oxide synthase ( iNOS ) , which leads to production of NO [14] , [15] , which in turn inhibits chlamydial growth [14] , [16] , [17] . Molecular motifs derived from C . pneumoniae are detected by several pattern recognition receptors , especially Toll-like receptor 2 ( TLR2 ) and TLR4 [18] , [19] . TLR4 recognizes chlamydial components such as lipopolysaccharide ( LPS ) and heat shock protein 60 ( cHSP60 ) [20]–[24] , and the intact organism stimulates TLR2 and TLR4-mediated responses [25] , [26] . TLR-mediated signaling triggered by C . pneumoniae-derived molecules instigates development of an inflammatory innate immune responses and TLR/MyD88 signaling plays an important role in host responses against C . pneumoniae infection [18] , [19] . Studies from our laboratory indicate that MyD88-null mice with C . pneumoniae lung infections are unable to mount a sufficient early inflammatory response against the pathogen [18] . These mice show marked delays in recruiting PMNs , CD8+ and CD4+ T cells to the lungs , and fail to clear the pathogen , but then develop a severe , late-stage , and persistent inflammation characterized by increased IL-1β and IFN-γ production that leads to increased mortality [18] . In contrast , TLR4−/− , TLR2−/− , and WT mice—all of which can detect C . pneumoniae and can signal normally via MyD88 , readily recovered from the infection and cleared bacteria normally , indicating that MyD88 is essential to an effective defense , but that TLR2 and TLR4 can both detect the pathogen and are therefore redundant [18] , [19] . C . pneumoniae has a unique biphasic developmental cycle that occurs within the chlamydial inclusion , a membrane-bound vacuole that is trafficked to the peri-Golgi region , where it avoids fusion with lysosomes and destruction , and are able to replicate intracellularly [27] , [28] . Chlamydia-mediated vesicular trafficking events transform the inclusion into a compartment from which chlamydiae can acquire nutrients and interfere with multiple host cell functions [29] , [30] . While residing intracellularly , the pathogen presumably is not detected by the cell surface TLR2 and TLR4 receptors; hence , it is unclear how C . pneumoniae might be detected and held in check once it has been taken up by the cell . C . pneumoniae–infected macrophages can limit bacterial growth by expression of IFN-γ , which in turn is controlled by TLR4/MyD88-dependent pathway . However , since Chlamydia can also induce IFN-γ in the absence of TLR4/MyD88 signaling [31] , a potential role for TLR-independent and intracellular recognition receptors , such as the nucleotide oligomerization domain ( Nod ) proteins , has been suggested [31] . Nod proteins and their adaptor molecule Rip2 also known as RICK or CARDIAK are key components of a family of cytosolic innate immune pattern recognition receptors [32]–[36] . Nod1 and Nod2 recognize molecules in the cytoplasm that originate from bacteria , including peptidoglycan ( PGN ) , a component of bacterial cell walls , and the muramyl dipeptide ( MDP ) structure found in almost all bacteria [37] . Both Nod1 and Nod2 signal via the serine/threonine Rip2 kinase [34] , [38] , [39] . Once activated , Rip2 mediates activation of NF-κB and the subsequent production of inflammatory cytokines such as TNF-α and IL-6 [40]–[42] . Although some reports indicate that Nod/Rip2-mediated signaling does not induce IFN-γ [43] , other studies show that combined TLR and Nod/Rip2 signaling together can lead to IFN-γ expression [44] . In the present study we show that the Nod/Rip2 signaling pathway is essential to detect intracellular C . pneumoniae and direct subsequent innate immune host defenses and bacterial clearance in a mouse model of pneumonia , in addition to the well-established role of the TLR/MyD88 pathway . Rip2−/− mice infected with C . pneumoniae displayed an impaired cytokine and chemokine release such as IFN-γ , KC and MIP2 , and showed impaired iNOS mRNA expression and NO production , and delayed neutrophil recruitment , which led to delayed bacterial clearance , an intense late-stage and persistent lung inflammation and increased mortality .
Rip2−/− mice and WT controls were infected intratracheally with C . pneumoniae ( 1×106 IFU/mouse ) and evaluated for lung inflammation by histopathological analysis . Tissue sections were obtained at 3 , 5 , 14 , and 35 days after infection , fixation and histological staining ( H&E ) was performed , and sections were graded for degree of inflammation in blinded fashion as detailed in the Materials and Methods Section . As expected , C . pneumoniae–infected WT mice developed marked lung inflammation as expected by days 5 and 14 and cleared the inflammation and recovered to baseline by day 35 ( Figure 1A and 1B ) . However , Rip2−/− mice developed significantly greater lung inflammation than WT mice by day 5 , and day 14 , which persisted until the end of the study period at day 35 ( Figure 1A and 1B ) . Innate immune responses , and particularly IFN-γ plays an important role in host defense against acute infection and in establishment of persistence of C . pneumoniae [10] . We , therefore , determined the production of cytokines such as IL-6 , IL-12 p40 and IFN-γ levels in BALF and lung homogenates from infected Rip2−/− and WT mice on days 3 , 5 and 14 . Concentrations of IL-6 , IL-12p40 , and IFN-γ were significantly reduced in BALF of Rip2−/− mice at day 3 compared to WT mice ( Figure 1C ) . However , by day 5 and day 14 , IL-6 , IL-12p40 and IFN-γ concentrations in the BALF and lung homogenates from Rip2−/− mice were significantly increased and exceeded levels in WT mice ( Figure 1C ) . Thus , in addition to increased histopathological inflammation seen in Rip2−/− mice on days 5 and 14 and during the later stages , we observed an initial impaired and delayed kinetics in cytokine production in C . pneumoniae–infected Rip2−/− mice on day 3 , which was also followed by a significant increased in cytokine production in the lungs on days 5 and 14 ( Figure 1C ) . We next measured IL-6 and IFN-γ levels in the supernatant of infected bone marrow–derived macrophages ( BMDM ) and whole lung cells ex-vivo . C . pneumoniae infection-induced cytokine production ex-vivo ( IL-6 , and IFN-γ release ) were significantly impaired in both Rip2−/− macrophages and whole lung cells compared to WT macrophages and whole lung cells ( Figure S1A–S1D ) . Our data show impaired cytokine production in Rip2−/− mice infected with C . pneumoniae early on day 3 following infection but a significant reversal and increase in cytokines and more severe and persistent lung inflammation by day 5 and 14 compared to WT mice ( Figure 1 ) . We hypothesized that this more severe and persistent lung inflammation was due to an inability of Rip2−/− mice to clear bacteria , which would then be expected to continue to provoke inflammation and cause the delayed increase in cytokine production . To test this hypothesis , we performed quantitative bacterial cultures in the lungs of mice at days 3 , 5 , and 14 post-infection . As anticipated , we observed significantly higher numbers of C . pneumoniae IFU in the lungs of Rip2−/− mice on days 5 and 14 compared to WT mice ( Figure 2A ) . This could not be explained by higher baseline load of bacteria in Rip2−/− mice , since on day 3 , bacterial numbers in lungs were similar between WT and Rip2-deficient mice ( Figure 2A ) . Consistent with the bacterial clearance data , virtually all WT mice survived the infectious challenge , while Rip2−/− mice had significantly increased mortality , and less than half the Rip2−/− mice survived until the end of the experiment at day 35 ( Figure 2B ) . Furthermore , the lungs from the Rip2−/− mice that succumbed to infection harbored an abundance of C . pneumoniae ( data not shown ) , while those who survived cleared the bacteria but still manifested chronic lung inflammation at day 35 ( Figure 1B ) . Collectively then , these data indicate that: 1 ) Rip2 importantly contributes to clearance of C . pneumoniae from the lungs; and 2 ) in the absence of Rip2 , severe lung inflammation occurs and persists , but fails to effectively combat the infection . Polymorphonuclear neutrophils ( PMN ) are crucial for innate host defense against bacteria and fungi . We previously reported that MyD88−/− mice infected with C . pneumoniae fails to recruit PMN into the lungs during early and late stages of the infection [18] . To investigate the PMN recruitment in Rip2-deficient mice , we infected Rip2−/− and WT mice with C . pneumoniae intratracheally , and compared total cells and PMN in BALF on day 3 and 5 following infection . Both PMN and total BALF cells in Rip2−/− mice were significantly lower compared to WT mice on day 3 following infection ( Figure 3A and 3B ) . However , by days 5 and day 14 post-infection , PMN as well as total BALF cell counts in Rip2−/− mice increased markedly , and were significantly higher than in WT mice ( Figure 3A and 3B ) . Assessment of neutrophil recruitment to the lung by flow cytometric analysis demonstrated similar results ( Figure 3C ) . The percentage of neutrophils ( defined by Gr1+ CD11b+ cells ) in the lungs of Rip2−/− mice were reduced on day 3 of infection , but increased thereafter , and by days 5 and 14 , significantly exceeded the neutrophil percentage of lung cells in WT mice ( Figure 3C ) . We next sought to examine whether the chemokines associated with neutrophil recruitment in the lungs were also affected in the Rip2-deficient mice . Rip2−/− mice showed significantly lower concentrations of KC and MIP-2 in both BALF and lung homogenates compared with WT mice on day 3 after infection ( Figure 3D ) . However , both KC and MIP-2 levels in BALF and lung homogenates significantly increased in Rip2−/− mice compared to WT mice by day 5 ( Figure 3D ) . Collectively , these data indicate that Rip2 plays an important role in early cytokine and chemokine production and neutrophil recruitment to the lungs during the initial days after C . pneumoniae infection , and Rip2-deficiency leads to delayed bacterial clearance , which is followed by an exaggerated secondary response consisting of increased cytokine and chemokine expression , PMN recruitment , prolonged , severe histopathological inflammation in the lungs , and increased mortality . Alveolar epithelial cells are the main cells infected in lung infection model [45] , but C . pneumoniae also infects different cell types including macrophages , dendritic cells , endothelial cells , and PMNs [7] . To determine which cells in the lungs are infected by C . pneumoniae , we analyzed infected cell profiles by flow cytometry . C . pneumoniae was predominantly found in macrophages and neutrophils , but also in alveolar epithelial cells in the infected lungs ( Figure 4A and 4B ) . Interestingly , in Rip2−/− mice , the number of neutrophils that contained C . pneumoniae was significantly increased compared to that in WT mice ( Figure 4A ) . To address whether more bacteria are in Rip2−/− macrophages and neutrophils , we analyzed mean fluorescence intensity ( MFI ) per cells , which corresponds to relative bacterial number ( Figure 4C ) . We observed a shifted histogram in Rip2−/− neutrophils . These data revealed that neutrophils are likely the main site of Chlamydial replication in lungs at day 5 after infection in Rip2−/− mice . We hypothesized that a bactericidal factor produced by immune effector cells might be responsible for the failure of Rip2−/− mice to clear bacteria . NO produced after cell activation by IFN-γ is important for killing or inhibiting growth of microorganisms [46] . Both IFN-γ and iNOS play major roles in host resistance to chlamydial infection [10] . We therefore assessed the levels of iNOS in the lungs following C . pneumoniae infection . Rip2−/− mice demonstrated significantly impaired iNOS mRNA expression compared to WT mice from day 0 until day 5 in total lung cells examined ex vivo ( Figure 5A ) . In addition , bone marrow-derived macrophages obtained from Rip2−/− mice , showed significantly diminished NO production following in vitro infection with C . pneumoniae compared to WT macrophages ( Figure 5B ) . These results suggest that NO production plays a role in killing and clearance of C . pneumoniae , and also suggest that Rip2 signaling contributes to NO production in response to C . pneumoniae infection . Consistent with this interpretation , C . pneumoniae growth was significantly increased in WT macrophages in the presence of an iNOS inhibitor ( L-NMMA ) compared to control cells treated with an inactive form of the inhibitor ( D-NMMA ) ( Figure 5C ) . In contrast , C . pneumoniae growth was not affected by treatment with the iNOS inhibitor in Rip2−/− macrophages ( Figure 5C ) . Collectively , these results suggest that Rip2-deficient mice have impaired iNOS expression and NO production in response to C . pneumoniae infection , which likely contribute to the host immune response defect and delayed bacterial clearance from the lungs of Rip2−/− mice . Our results thus far indicate that the Rip2−/− mice display an impaired host defenses , delayed bacterial clearance , and increased mortality following C . pneumoniae lung infection . Since Rip2 is utilized by both Nod1 and Nod2 , we next wished to determine the role of these upstream receptors in C . pneumoniae infection . Nod1 was shown to play a role in C . pneumoniae-mediated activation of human endothelial cells in vitro [47] , But it is unclear which Nod receptors detect C . pneumoniae in macrophages and during in vivo infection . Nod1 is ubiquitously expressed in mammalian cells , but the expression of Nod2 is mainly restricted to primary antigen-presenting cells and epithelial cells , and Nod2 is not expressed in endothelial cells [48] . Furthermore , our data ( Figure 4C ) indicates that C . pneumoniae mainly replicates in macrophages and neutrophils . To determine which Nod receptor recognizes intracellular C . pneumoniae in macrophages , we infected Nod1−/− or Nod2−/− BMDM with live C . pneumoniae and measured KC and NO levels in the supernatant . Nod1−/− and Nod2−/− macrophages produced significantly diminished KC and NO ( Figure 6A and 6B ) . Consistent with this decreased NO production , bacterial viability was significantly higher in both Nod1−/− and Nod2−/− macrophages in vitro ( Figure 6C ) . To determine if greater bacterial viability in the absence of Nod1 and Nod2 also occurred in vivo , we infected Nod1−/− and Nod2−/− mice with C . pneumoniae and examined bacterial clearance in the lungs . In agreement with the in vitro data , both Nod1−/− and Nod2−/− mice displayed delayed pulmonary bacterial clearance compared to WT controls , as reflected by significantly higher bacterial counts in Nod1−/− and Nod2−/− mice at 5 days post-infection ( Figure 6D ) . These results are consistent with the conclusion that intracellular C . pneumoniae is recognized by both Nod1 and Nod2 in macrophages , and that signaling emanating from both Nod1 and Nod2 significantly contributes in host defenses against C . pneumoniae lung infection , at least in part by regulating production of NO and inflammatory cytokines and chemokines such as IL-12 p40 , IFN-γ , KC and MIP2 . Based upon data in the previous section that showed involvement of macrophages and PMNs in the lungs , we hypothesized that the Nod/Rip2 signaling pathway in bone marrow ( BM ) -derived cells rather than non-hematopoietic stromal cells was primarily responsible for innate immune host responses and clearance of C . pneumoniae from the lungs . To test this notion , we generated chimeric mice using donor marrow from WT or Rip2−/− mice ( Figure S2 ) , then infected the mice intratracheally with C . pneumoniae . Five days after infection , lungs were harvested and quantitative bacterial counts were determined . WT recipient mice that were transplanted with Rip2−/− BM displayed significantly higher bacterial load in the lungs compared to control WT mice transplanted with WT BM ( Figure 7A ) . Conversely , Rip2−/− recipient mice that had been transplanted with WT BM displayed lower bacterial counts than control Rip2−/− mice that had received Rip2−/− BM ( Figure 7A ) . In all chimeric mice , we observed generally higher bacterial titers observed in the lungs , most likely due to inherently increased susceptibility secondary to the irradiation procedure itself , as has been previously reported by other investigators [49] . These data indicate that Rip2 in BM-derived cells primarily mediates host defenses against pulmonary C . pneumoniae infection . However , it is possible that airway epithelial cells also contribute and play a role in C . pneumoniae detection in the lung . In order to further elucidate the primary role of macrophages in C . pneumoniae infection , we performed intratracheal adoptive transfer of WT or Rip2−/− BMDM , simultaneously with C . pneumoniae infection ( i . e . macrophages mixed with bacteria ) , and then determined the effect on local bacterial replication in lungs of WT or Rip2−/− mice . As anticipated , adoptive transfer of Rip2−/− macrophages plus C . pneumoniae into WT mice resulted in significantly higher bacterial counts compared to WT macrophages plus C . pneumoniae transferred into WT mice ( Figure 7B ) . However , WT macrophages plus C . pneumoniae adoptively transferred into Rip2−/− mice rescued the Rip2 phenotype , i . e . restored bacterial clearance ( Figure 7B ) and neutrophil recruitment in the lungs ( Figure S3 ) , indistinguishable from control WT mice that received WT macrophages plus C . pneumoniae . We did not observe a defect in phagocytosis in Rip2−/− macrophages using labeled C . pneumoniae ( Figure S4 ) [50] . Taken together , these findings indicate that the Nod/Rip2 signaling pathway in BM-derived cells play a dominant role in bacterial clearance of C . pneumoniae from the lungs .
Here we show that cytoplasmic Nod proteins are importantly involved in generating innate immune defenses against intracellular C . pneumoniae . We found that deletion of Rip2 , that is essential for both Nod1- and Nod2-mediated signaling , delays neutrophil recruitment to the lungs , and suppresses expression of chemokines and cytokines that are essential to generate an effective host defense . Although an inflammatory innate immune response was delayed , by day 5 Rip2−/− mice infected with C . pneumoniae developed a more severe inflammation that persisted longer compared to WT mice , but nevertheless failed to clear the pathogen , and most infected Rip2−/− mice ultimately succumbed to the infection . The inability of Rip2−/− mice to eradicate the pathogen despite robust inflammation was associated with delayed kinetics of IL-12p40 and IFN-γ production and suppressed iNOS expression and NO production , all of which are critically important elements in the innate immune armamentarium [9] , [11] . Experiments with bone marrow-derived macrophages demonstrated that both Nod1 and Nod2 were involved in sensing intracellular C . pneumoniae . Results from experiments with bone marrow chimeric mice confirmed that cells derived from hematopoietic lineages rather than resident stromal cells were essential in Nod/Rip2-mediated defenses against C . pneumoniae . This conclusion was corroborated by adoptive transfer of WT or Rip2−/− macrophages directly into the airways of infected mice . Collectively , our data demonstrate that proper functioning of the Nod/Rip2 cytoplasmic innate immune detection system critically determines whether the host can effectively resist and eradicate an infectious challenge to the lungs from C . pneumoniae . Our results underscore previous reports that increasingly emphasize the central role Nod/Rip2 signaling can play in defending the host against intracellular invasion . Nod proteins mediate host defense against a variety of both Gram negative and Gram positive bacteria . For example , Nod2 senses the PGN produced by Staphylococcus aureus , and Rip2 limits S . aureus growth in macrophages [51] . Also , Nod2 detects the MDP structure found in almost all bacteria [37] . Nod1 is required for IkK and NF-κB activation in human colon epithelial cells infected with E . coli [52] , participates in KC induction and impacts bacterial viability to Pseudomonas aeruginosa in mouse embryonic fibroblasts [53] . Nod2 triggers cytokine production by DCs in response to live M . tuberculosis , but is not essential to control M . tuberculosis airway infection [54] . While Rip2 and Nod1 deficiency increases susceptibility to Listeria monocytogenes [32] and Helicobacter pylori [55] respectively , neither Nod1 nor Rip2 deficiency had any significant effect on Chlamydia muridarum vaginal infection [56] . Hence , available data indicates that Nod1 and Nod2 selectively interact with specific pathogens and can play a critical role in host defenses , highlighted by increased susceptibility to specific pathogens in mice lacking these intracellular receptors . Accordingly , our data now indicate that Nod/Rip2 signaling is also essential in successfully combating C . pneumoniae lung infection . Indeed , in the absence of Nod/Rip2 signaling , lung infection with C . pneumoniae proved fatal to the majority of mice . Clearly , one reason for the increased susceptibility of the Rip2−/− mice to C . pneumoniae lung infection may be the inability of Rip2−/− mice to rapidly recruit neutrophils to the site of infection . Delayed neutrophil recruitment in Rip2−/− mice appears closely linked to lack of sufficient chemokine and cytokine expression compared to infected WT mice . However , by days 5 and 14 post-infection , the percentage of neutrophils in the lungs of Rip2−/− mice increased and significantly exceeded those in WT mice . Evidently , this delayed early response to the infectious challenge allows the bacteria to gain the upper hand , but despite the significant increase in PMN numbers on days 5 and 14 post-infection , the Rip2−/− mice showed delayed bacterial clearance from the lungs and increased mortality . Interestingly , in Rip2−/− mice , the number of neutrophils in the lungs that contained C . pneumoniae was significantly increased compared to that in WT mice . Several obligate intracellular microbial pathogens develop mechanisms to evade destruction upon ingestion by PMN [57] , [58] . Indeed , C . pneumoniae can infect and replicate in PMNs and these cells in turn can enhance replication and Chlamydial burden during infection [59] . However , the main reason Rip2−/− mice cannot clear the infection and most often succumb to the disease appears to be more closely tied to our observation that expression of IL-12p40 , IFN-γ , iNOS and NO are suppressed in the absence of Rip2−/− . These data are consistent with previous reports , which similarly indicate that IL-12p40 [9] , IFN-γ [9] , [60] and iNOS [11] , [61] are essential for effective host resistance against C . pneumoniae infection . Several studies showed that IFN-γ and IL-12 both play an important role in the innate control of C . pneumoniae infection [9] , [11] IL-12 participates in resistance to C . pneumoniae , likely by enhancing IFN-γ mRNA [9] . In turn , IFN-γ produced by innate cells increases iNOS expression and NO release and controls the intracellular growth of C . pneumoniae [14] , [16] , [17] . Increased susceptibility of IFN-γR−/− mice is associated with diminished levels of iNOS mRNA accumulation in lungs , and iNOS−/− mice also show higher sensitivity to C . pneumoniae infection [11] . However , IFN-γR−/− mice shows even greater sensitivity to C . pneumoniae infection compare to iNOS−/− mice , suggesting the presence of both iNOS-dependent and -independent IFN-γ-mediated effector mechanisms [11] . IFN-γ can be produced by cells from both the innate and acquired immune system . The susceptibility of IFN-γR−/− mice largely exceeds that of RAG-1−/− mice , suggesting an important role for non-T cell-mediated IFN-γ-producing cells in the host resistance against C . pneumoniae infection [11] . Several studies show that , besides NK and T cells , myeloid cells such as macrophages , DCs and neutrophils can also express IFN-γ [62] . Previous studies have shown that MDP induced NO production in macrophages [63] , [64] . Several other reports also suggest a link between Nod and NO production [64] . Our results indicating that Nod/Rip2 signaling stimulates iNOS expression and NO production suggest that one of the reasons why C . pneumoniae lung infection proves lethal to most Rip2−/− mice is because they fail to generate an effective NO-mediated defense by immune effector cells , and thus cannot eradicate the pathogen . In addition to innate immunity , adaptive immune responses may also directly or indirectly diminish the levels of IFN-γ and IL-12 mRNA early after infection and thus may alter the quality of the protective host immune responses . A protective role for CD8 T cells is shown by the higher sensitivity and enhances severity of infection in CD8−/− mice [11] . In MyD88−/− mice we observed a delay in recruitment of CD4 and CD8 T cells into the lungs [18] , while in the current study with Rip2−/− mice we observed primarily a delay CD4 T cell recruitment initially followed by significant increases in the presence of both CD4 and CD8 T cells later on day14 ( Figure S5 ) . Our data with bone marrow chimeric mice clearly demonstrate that the cells responsible for Nod/Rip2-dependent defense against C . pneumoniae are hematopoietic in origin , and are not resident stromal cells . Additionally , adoptive transfer of WT macrophages was able to rescue the bacterial clearance defect in Rip2−/− mice . While our data do not completely rule out a potential role for other cell types during C . pneumoniae infection , the bactericidal effects of the Nod/Rip2 pathway appear to be predominantly of bone marrow origin with macrophages playing the largest role . C . pneumoniae-induced IL-12p40 production in vivo involves both MyD88-dependent and MyD88-independent pathways [59] , suggesting a TLR-independent , but Nod-dependent mechanism of recognition and activation . Indeed , recent studies suggest that in endothelial cells , Nod1 plays an important role in triggering C . pneumoniae-mediated inflammatory responses [47] , and that Nod1 is involved in NF-κB activation by Chlamydia in epithelial cell lines [56] . Furthermore , a recent study by Buchholz et . al . concluded that C . trachomatis induced IL-8 responses are dependent on Nod1 and Rip2 signaling in Hela cells [65] . Our in vivo data showing that C . pneumoniae induced chemokine production in the lungs depends on Rip2 signaling is consistent with the in vitro observations by Buchholz et . al . These findings suggest that PGN fragments are synthesized by chlamydiae and are recognized by the host innate immune system . The genome sequence revealed that Chlamydophila is actually equipped with a full complement of PGN synthesis genes [66] . Chlamydia is sensitive to antibiotics like penicillin that inhibit PGN synthesis [67] , [68] , but clear-cut biochemical evidence for the synthesis of PGN in chlamydiae is missing [69] , [70] . A recent study revealed the biochemical capacity of C . trachomatis to synthesize m-DAP and that the m-DAP synthesis genes are expressed as early as 8h after infection [71] . This paradox , known as the ‘chlamydial anomaly’ is still being debated in the light of genomic information [72] . However , prior studies and our current data suggest that chlamydial PGN released by bacteria must make their way across the inclusion membrane into the cytosol . One potential mechanism by which this could occur is through the proposed type III secretion system [73] . A similar mechanism of type IV secretion has been proposed for Nod1 signaling in H . pylori infection [55] . While the exact ligand ( s ) of C . pneumoniae detected by the Nod are yet to be identified , our data clearly indicate that both Nod1 and Nod2 recognize C . pneumoniae and play an essential role in host defenses against this microorganism . Our data differ from those obtained in experimental infections with C . trachomatis or C . muridarum genital tract infection , where Nod1 deficiency had no significant effect on the efficiency of infection , or pathology in vaginally infected mice , while Rip2-deficient mice had only slightly increased bacterial load and delayed bacterial clearance and mildly increased oviduct inflammation [56] . Such differences are not surprising , as the two organisms display only 5 and 10% homology at the DNA and protein levels , respectively , as also reflected in the different pathobiologies they cause [74] . In summary , we demonstrate that the Nod cytosolic pattern recognition receptors are essential for mounting an adequate defense against C . pneumoniae , that Nod stimulate chemokine and cytokine production and neutrophil recruitment in the early phase of infection , and that the cells responsible for the effects of Nod are bone marrow-derived cells , not stromal cells . Furthermore , we show that Nods stimulate IL12-p40 , IFN-γ , iNOS and NO expression , and that these factors are key for surviving the infectious challenge . Since the TLR/MyD88 pathway is also critically involved in detecting and eradicating C . pneumoniae , our data highlight an emerging theme in host defenses: that divergent pattern recognition receptors that are seemingly unrelated and expressed in distinct compartments can nevertheless direct cooperative responses that successfully combat invasion by common pathogens such as C . pneumoniae . Coordinated and sequential activation of TLR and Nod signaling pathways may be necessary for efficient immune responses and host defenses against C . pneumoniae . While TLRs might be important for initial activation upon Chlamydophila contact , it is likely that Nod proteins play a role in the sequential and intracellularly triggered prolonged activation of target cells by intracellular Chlamydophila .
Rip2−/− mice , backcrossed ten generation to C57BL/6 , were kindly provided by Dr . Genhong Cheng ( University of California at Los Angeles , Los Angeles , CA , USA ) . C57BL/6 mice and Nod2−/− mice were purchased from Jackson Laboratory . Nod1−/− mice were kindly provided by Dr . Jeffrey Weiser ( University of Pennsylvania , Philadelphia , PA , USA ) . Mice were maintained under specific pathogen-free conditions , and were used at 8–12 weeks of age . All experiments were done according to Cedars-Sinai Medical Center Institutional Animal Care and Use Committee guidelines . C . pneumoniae CM-1 ( ATCC , Manassa , VA ) was propagated in HEp-2 cells as previously described [18] . HEp-2 cells and C . pneumoniae stocks were determined to be free of Mycoplasma contamination by PCR . Mice were intratracheally infected with C . pneumoniae by inoculating 100 µl of PBS containing 1×106 IFU of the microorganism . Bronchoalveolar lavage fluid ( BALF ) was collected with 0 . 5 ml of PBS containing 2mM EDTA . The lavage fluid was centrifuged , and the supernatant was used for chemokine and cytokine measurements . The pellet placed on glass slides , and stained by modified Wright-Giemsa staining ( Diff-Quick; Fisher Scientific , Pittsburgh , PA , USA ) to determine leukocyte subtypes based on their cellular and nuclear morphology . Lungs were homogenized with 1ml of sucrose-phosphate-glutamate medium and stored at −80°C . To quantify C . pneumoniae progeny , HEp2 cells were inoculated with lung specimens or cell lysates as previously described [75] . Briefly HEp2 cells were infected with diluted lung homogenates or infected cell lysates . Cultures were centrifuged for 1h at 800× g , fed with RPMI1640 in the presence of cycloheximide ( 1 µg/ml ) , and incubated for 72h . Thereafter , Cells were washed with PBS , fixed with methanol for 5 min at room temperature and stained with FITC-conjugated Chlamydia genus-specific mAb ( Pathfinder Chlamydia Culture Confirmation System; BIO-RAD , Hercules , CA , USA ) according the manufacturer protocol . Inclusion bodies were counted by fluorescence microscopy . Lungs were fixed in formalin buffer , paraffin-embedded , and hematoxylin and eosin-stained sections were scored by a trained pathologist blinded to the genotypes as previously described [18] . Briefly , the degree of inflammation was assigned an arbitrary score of 0 ( normal = no inflammation ) , 1 ( minimal = perivascular , peribronchial , or patchy interstitial inflammation involving less than 10% of lung volume ) , 2 ( mild = perivascular , peribronchial , or patchy interstitial inflammation involving 10–20% of lung volume ) , 3 ( moderate = perivascular , peribronchial , patchy interstitial , or diffuse inflammation involving 20–50% of lung volume ) , and 4 ( severe = diffuse inflammation involving more than 50% of lung volume ) . The chemokine and cytokine concentrations in the BALF , lung homogenates or culture supernatant were determined using by Duoset Mouse KC , MIP-2 ( R&D systems , Minneapolis , MN , USA ) , OptiEIA Mouse IL-6 ELISA Set ( BD Biosciences , San Jose , CA , USA ) and Mouse IFN-γ ELISA , Mouse IL-12 p40 ELISA ( eBioscience , San Diego , CA , USA ) . The assays were performed as described manufacturer protocol . The lymphocytic makeup in the lungs after infection were analyzed by flow cytometry of lung homogenates . Briefly , lymphocytes were isolated by digesting the lung tissue at 37°C for 1h with HANKS' containing 100 µg/ml Blenzyme ( Roche Diagnostics , Indianapolis , IN , USA ) and 50 units/ml DNase I ( Roche Diagnostics ) and filtering through a 70 µm cell strainer ( BD Biosciences ) . Erythrocytes were depleted by lysis buffer before staining . Isolated single cells were stained with following specific mAbs; CD16/32 ( clone 93 ) , Gr1 ( clone RB6-8C5 ) , CD11b ( clone M1/70 ) , F4/80 ( clone BM8 ) , CD11c ( clone HL3 ) , CD45 ( clone 30-F11 ) , CD4 ( clone RM4-5 ) and CD8 ( clone 53-6 . 7 ) were purchased from eBioscience as direct conjugates to FITC , PE or PECy5 . Anti SP-C polyclonal Ab and PEcy5-conjugated donkey anti-Goat IgG F ( ab' ) were used for Alveolar type II epithelial ( ATII ) cell staining ( Santa Cruz Biotechnology , Santa Cruz , CA , USA ) . Cells were identified based on expression of following antigens: pulmonary macrophages ( F4/80+ and CD11c+ ) , DC ( F4/80− and CD11c+ ) , Neutrophils ( Gr1+ and CD11b+ ) , ATII cells ( SP-C+ , CD45- and CD16/32- ) , T cells ( CD3+ ) , B cells ( CD19+ ) . For intracellular Chlamydophila staining , cells were permeabilized using Cytofix/Cytoperm kit ( BD Biosciences ) and stained with FITC-conjugated anti-Chlamydia LPS mAb ( Accurate Chemical and Scientific Corporation , Westbury , NY , USA ) . Flow cytometric analysis was performed by FACScan flow cytometer ( BD Biosciences ) and the data was analyzed by Summit ( Dako , Carpinteria , CA , USA ) . Total RNA was extracted from homogenized lung tissues by RNeasey mini kit ( QIAGEN , Valencia , CA , USA ) following the manufacturer's protocol . Total RNA preparations were subjected to reverse transcriptase-polymerase chain reaction analysis by Total cDNA was generated using the Omniscript cDNA synthesis kit ( Qiagen ) , PCR analysis was performed using specific primers for mouse iNOS ( sense: 5′-TGG GAA TGG AGA CTG TCC CAG-3′:antisense: 5′-GGG ATC TGA ATG TGA TGT TTG-3′ ) , 1min at 94°C , 1 min at 58°C and 2 min at 68°C . Amplification of GAPDH served as a control . Femora and tibiae of mice were rinsed with cell culture medium . Bone marrow cells were treated with red blood lysis buffer ( eBiosciences ) , cultured in RPMI1640 medium containing 10% FBS and 10 ng/ml M-CSF ( R&D system ) . Medium changed at day 3 and day 6 . BMDM were harvested at day 9 and exposed to C . pneumoniae by centrifugation at 500× g for 30 min . Nitrite levels in the culture supernatant were determined using the colorimetric Griess reaction ( Sigma , St . Louis , MO , USA ) . Absorbance was measured with a plate reader at 540 nm . The concentration of NO2− was determined from standard curves constructed with serial concentrations of NaNO2 . Recipient WT ( Ly5 . 1 ) , WT ( Ly5 . 2 ) and Rip2−/− ( Ly5 . 2 ) mice were lethally γ-irradiated with 950 rads using a 137Cs γ-source and were reconstituted intravenously with 5×106 BM cells derived from respective donors 2–3h later . All mice were placed on Baytril ( Bayer HealthCare LLC , Shawnee Mission , KS , USA ) for 2 weeks following irradiation . 6–7 weeks after engraftment , mice were tested by FACS analysis with FITC-conjugated Ly5 . 2 Ab ( clone 104 , eBiosciences ) and PE-conjugated Ly5 . 1 Ab ( clone A20 , eBiosciences ) staining for chimerism . Data are reported as mean values±S . D . Statistical significance was evaluated by Student's t test . In the case of survival study , Statistical significance was evaluated by Fisher's exact test . For multiple comparison test , Statistical significance was evaluated by one way ANOVA with Tukey's post-hoc test . | Chlamydophila pneumoniae ( C . pneumoniae ) is a common intracellular parasite that causes lung infections and contributes to several diseases characterized by chronic inflammation . Toll-like receptors expressed on the cell surface detect C . pneumoniae and mount a vigorous defense , but it is not known how the cell defends itself once the pathogen has taken up residence as a parasite . We reasoned that cytosolic pattern recognition receptors called Nods ( nucleotide oligomerization domain ) that detect microbes that gain entry into the cell might be involved . Using mice genetically deficient in Nod1 and Nod2 or their common downstream adaptor ( Rip2 ) , we show that in lung infection , Nod proteins are indeed essential in directing a defense against C . pneumoniae . Mice with defective Nod/Rip2-dependent signaling exhibited delayed recruitment of neutrophils , blunted production of pro-inflammatory cytokines and chemokines , and evidence of defective iNOS expression and NO production . These impaired responses led to delayed clearance of bacteria , intense persistent lung inflammation , and increased mortality . By performing bone marrow transplantation experiments and direct transfer of cells into the lungs of mice , we demonstrated that intact Nod-dependent signaling in bone marrow–derived cells was critical in the defense against C . pneumoniae . Our results indicate that Nod proteins also play an important role in host defense against C . pneumoniae . Coordinated and sequential activation of TLR and Nod signaling pathways may be necessary for an efficient immune response and host defense against C . pneumoniae . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"microbiology/immunity",
"to",
"infections",
"microbiology/innate",
"immunity",
"immunology/innate",
"immunity",
"pathology/immunology",
"infectious",
"diseases/respiratory",
"infections",
"infectious",
"diseases/bacterial... | 2009 | The NOD/RIP2 Pathway Is Essential for Host Defenses Against Chlamydophila pneumoniae Lung Infection |
Changes in the spatial positioning of genes within the mammalian nucleus have been associated with transcriptional differences and thus have been hypothesized as a mode of regulation . In particular , the localization of genes to the nuclear and nucleolar peripheries is associated with transcriptional repression . However , the mechanistic basis , including the pertinent cis- elements , for such associations remains largely unknown . Here , we provide evidence that demonstrates a 119 bp 5S rDNA can influence nucleolar association in mammals . We found that integration of transgenes with 5S rDNA significantly increases the association of the host region with the nucleolus , and their degree of association correlates strongly with repression of a linked reporter gene . We further show that this mechanism may be functional in endogenous contexts: pseudogenes derived from 5S rDNA show biased conservation of their internal transcription factor binding sites and , in some cases , are frequently associated with the nucleolus . These results demonstrate that 5S rDNA sequence can significantly contribute to the positioning of a locus and suggest a novel , endogenous mechanism for nuclear organization in mammals .
The organization of DNA within mammalian nuclei is considered nonrandom [1] . A number of characteristics have been proposed to influence the position of a gene or chromosomal region within the nucleus , including gene density and transcriptional activity [2] . However , the parameters that drive nuclear organization are likely complex and remain largely enigmatic . Significant proportions of mammalian genomes are comprised of noncoding , repetitive elements , many of which are derived from RNA polymerase III ( pol III ) transcripts . An increasing number of examples have suggested diverse roles for repetitive elements in modulating transcription of neighboring protein-coding genes transcribed by RNA polymerase II ( pol II ) [3] , [4] , [5] , [6] . In yeast , binding sites for the pol III transcription factor complex , TFIIIC , play a significant role in chromatin structure and nuclear organization: tRNA genes and tRNA-like sequences function as chromatin barriers to prevent the spread of heterochromatin , while in other contexts these elements cluster together often at the nuclear and nucleolar peripheries [7] , [8] . This latter phenomenon typically results in silencing of nearby pol II-transcribed genes [9] . Moreover , just as pol II genes are thought to cluster in transcription ‘factories’ [10] , active pol III also forms distinct foci in mammalian nuclei that contain a number of active pol III genes [11] . Since most pol III transcribed genes , including those of repetitive elements , carry internal promoters , they could confer intrinsic structural and regulatory properties to the surrounding genomic sequence upon insertion . Given their widespread and nonuniform distribution in mammalian genomes through repetitive elements , pol III promoters may have significant influence on chromatin structure . Furthermore , binding sites for pol III transcription factors within these elements may be under positive selection if beneficial for host genome fitness . To test these hypotheses , we focused on 5S rRNA genes ( Figure 1A ) , which have long been known to possess unique qualities with regard to chromatin structure . We use a number of complimentary approaches to demonstrate that ectopic 5S rDNA sequence can mediate nucleolar association of a genomic region , with significant effects on local transcription . We also provide evidence that this mechanism may be active in endogenous contexts in the mouse genome: psuedogenes that are derived from 5S rDNA show preferential conservation of internal transcription factor binding sites can be bound by TFIIIC and localize to the nucleolar periphery .
A well-known nucleosome positioning sequence , 5S rDNA genes ( endogenously present as multi-copy arrays in most eukaryotic genomes ) have been observed to form large chromatin loops in Xenopus and mammalian systems [12] , [13] . In agreement with observations in other eukaryotes , and recently published descriptions of chromatin associated with nucleoli in human cells [14] , [15] , [16] , we found the mouse 5S rDNA gene array ( located on the distal end of chromosome 8 ) associated with the nucleolar periphery in ∼40% of mouse embryonic stem ( ES ) cells ( Figure S1A ) . If localization to the nucleolar periphery is an intrinsic quality of the 5S rRNA genes , then de novo insertion of these sequences into new genomic contexts should recapitulate this phenomenon . To study the effect of 5S rDNA sequence on sub-nuclear localization , we generated ES cell lines with stable , multicopy insertions of a reporter construct containing a single 5S rRNA gene ( Tg5S ) ( Figure 1B ) . To determine whether transgenes with 5S sequence would be found at the nucleolar periphery , we then assessed localization of the stable transgenes by DNA FISH with a probe for the vector backbone relative to immunofluoresence against Nucleolin , a marker for the nucleolus [17] . In support of our hypothesis , we observed significantly more frequent localization to the nucleolar periphery of Tg5S ( 75% ) compared with empty vector controls ( Tg0 , 31% , p = 8×10−4 , Figure 1C , Figure S1B–S1D ) . Strikingly , several lines showed nearly constitutive association of the Tg5S signal with the nucleolus . This was not simply a reflection of copy number , as this pattern of localization was observed in both high- and low-copy Tg5S lines ( Figure S1E , R2 = 0 . 087 ) . Furthermore , association of Tg5S was higher than that of the 5S rDNA array ( ∼40% ) . This could be due to a dominant localization pattern imparted by Tg5S even at low copy , or additional forces acting to constrain localization of the endogenous 5S rDNA locus . We observed very little co-localization of Tg5S arrays and the 5S rDNA cluster ( <1% , data not shown ) , demonstrating that these loci do not occupy the same compartment in the nucleoplasm . However , we found that the structural and functional integrity of the nucleolus was essential for localization through 5S rDNA . Reorganization of nucleolar components , through pharmocological inhibition of RNA polymerase I activity , resulted in a significant decrease of both Tg5S and 5S rDNA association with the nucleolus ( Figure S2 ) . The nucleolar periphery has typically been thought of as a transcriptionally quiescent compartment , often associated with examples of constitutive [18] , [19] , [20] and facultative [21] , [22] , [23] heterochromatin . To study the effects of nucleolar association through the 5S rDNA mechanism on pol II transcription , we quantified mRNA levels of a reporter gene present on the vector: the Thymidine kinase ( Tk ) gene driven by the mouse Pgk1 promoter ( Figure 1B ) . Tk mRNA levels , when normalized for copy number , are significantly decreased in Tg5S lines compared with Tg0 lines ( 4 . 68±2 . 22 and 8 . 09±1 . 55 arbitrary units , respectively; p = 6×10−3 , Figure 1D , Figure S3 ) . Interestingly , Tk mRNA levels show a strong negative correlation with nucleolar association: lines with the most frequent association had the lowest normalized expression ( Figure 1E , R2 = 0 . 664 ) . This relationship suggests that perinucleolar targeting of transgenes via the 5S rDNA sequence has inhibitory effects on pol II transcription . The efficiency of nucleolar localization and transcriptional repression observed by Tg5S may be related to its ability to recruit the pol III transcriptional machinery . In yeast , the regulatory capacity of tRNA and tRNA-like sequences is dependent upon the TFIIIC complex [14] . To determine whether the TFIIIC complex is associated with transgene-5S rDNA , we used chromatin immunoprecipitation ( ChIP ) for a subunit of TFIIIC , TFIIIC65 . We observed significant levels of TFIIIC65 association with transgene-5S rDNA , relative to the negative control ( the Ascl2 promoter ) , in three of four Tg5S lines analyzed ( Figure 2A ) . However , TFIIIC65 enrichment showed no clear correlation with localization ( Figure 2B ) , Tk mRNA levels ( Figure 2C ) , or copy number ( Figure 2D ) . These data suggest that while the TFIIIC complex may participate in the localization and transcriptional attenuation we have observed for the 5S transgenes , levels of TFIIIC65 alone are not sufficient to explain these phenomena . To determine whether specific histone modifications characterize the presence of a 5S rDNA , we surveyed the distribution of several modifications at various positions within the transgenes ( Figure 3A ) . We analyzed one mark of active chromatin ( H3K4me2 , Figure 3B , Figure S4A ) , one mark of constitutive heterochromatin ( H3K9me3 , Figure 3C , Figure S4B ) , and two marks of facultative heterochromatin ( H3K9me2 and H3K27me3 , Figure 3D , 3E , Figure S4C , S4D ) , in four Tg5S and two Tg0 cell lines . As expected , cell lines with higher expression of Tk ( Figure 3F ) had increased levels of H3K4me2 at the Tk gene . Intriguingly , all Tg5S lines were characterized by high levels of H3K9me3 near the 5S rDNA , rather than the Tk gene body or promoter . Both patterns were evident irrespective of TFIIIC65 enrichment to the transgene-5S rDNA ( Figure 3F ) . These observations suggest an association between the 5S rDNA sequence and the H3K9me3 modification . The frequent nucleolar association of 5S rDNA-containing transgenes suggests the capacity to direct localization of a genomic region to the nucleolar periphery . However , this observation may also reflect preferential integration of Tg5S into a chromosomal region neighboring the nucleolus in the parental cells , rather than a change in localization . To discriminate between these possibilities , we identified the insertion site for several Tg5S ES lines . We mapped the transgene insertion in Tg5S#9 to the pseudoautosomal region ( PAR ) of the X chromosome [24] ( Figure S5A ) . Since these ES cells are XY , we used the X-chromosome PAR of a line without a transgene insertion in this region as a control ( Tg5S#6 ) to assess localization changes relative to a homolgous , wild-type chromosome . The PAR with the transgene insertion was more frequently associated with the nucleolus ( 61% ) than a wild-type PAR ( 43% , p = 2×10−3 , Figure 4A , 4B ) . Although nucleolar association of the wt PAR was similar to that of the 5S rDNA locus ( 39% ) , this frequency increased significantly upon Tg5S insertion . Tg5S line #6 ( Tg5S#6 ) , contains an integration in the first intron of the silent RAR-related orphan receptor beta ( Rorb ) gene ( Figure S5B ) . The allele containing the transgene array was discernable by DNA FISH and always overlapped with the genomic probe ( Figure 4C ) . Nucleolar association was measured for both the wild type allele ( wt allele ) and the allele containing the Tg5S insertion ( tg allele ) . As a control , we measured localization of the Rorb alleles in Tg5S#9 , which does not have an insertion in this region . We detected significantly more DNA FISH signals for the tg allele associated with or internal to the nucleolus ( 68% ) than for the wt allele ( 52% ) in Tg5S#6 ( Figure 4D , p = 0 . 01 ) , or either allele in the control cell line ( 43% , p = 4×10−4 ) . The localization frequency of the wt allele in the Tg5S#6 was not significantly different from the alleles in the control line ( p = 0 . 5 ) . Interestingly , wt Rorb alleles were associated with the nucleolus significantly more frequently than the 5S rDNA locus ( chi-squared test , p = 5×10−9 ) . Together , our observations from two independent insertion events , in two very different genomic contexts , demonstrate that ectopic 5S rDNA can influence the position of a locus . Since localization by a Tg5S was associated with decreased transcriptional output of the Tk reporter gene , we hypothesized that the transgene insertion into the Rorb locus may similarly affect transcription of this gene . Rorb is not expressed in undifferentiated ES cells , therefore we differentiated the line with the Tg5S insertion in the Rorb gene ( Tg5S#6 ) along with Tg5S#9 , where the transgene insertion is not at the Rorb locus . Although activation of Rorb was variable between biological replicates , in each case Rorb expression was significantly reduced in Tg5S#6 ( Figure 4E ) . Intriguingly , average Rorb expression in Tg5S#6 was 60% of that in Tg5S#9 , suggesting that the presence of Tg5S at the Rorb locus has detrimental effects on its transcriptional activation . The mouse genome contains >110 5S rDNA genes annotated outside the array on chromosome 8 ( NCBI m37 mouse assembly , Table S1 , Figure S6A ) . However , these elements show low overall sequence conservation and no predicted structural similarity to bona fide 5S rDNA , and are therefore unlikely to be functional components of the large ribosomal subunit ( Figure S6B , S6C ) . Despite acquiring numerous mutations , a high proportion of these 5S pseudogenes retain perfect , or near-perfect , internal transcription factor binding sites ( Figure 1A ) . This conservation correlates poorly with overall similarity of the 5S pseudogenes to the 5S rDNA consensus ( R2 = 0 . 113 , Figure 5A ) , suggesting this is not simply due to recent insertion events , but rather indicative of differential selective pressure within the psuedogene . We found a subset 5S pseudogene loci associated with the nucleolus in E14 ES cells at a frequency comprable to that of the 5S rDNA locus ( Figure 5B , Figure S7 ) , further supporting a positional effect for this sequence . TFIIIC association with pseudogenes was not well correlated with localization: by ChIP , we observed high levels of TFIIIC65 enrichment at only one of two pseudogene loci frequently associated with the nucleolus ( Figure 5C ) . Therefore , if nucleolar association of these regions is mediated through 5S pseudogenes , then it may not require stable association of the TFIIIC complex , or perhaps involve altogether different mechanisms . Irrespective of the putative trans-factor , frequent nucleolar association of 5S pseudogenes further support a previously uncharacterized role for for these sequences as organizational cis-elements in the mammalian genome .
The relationship between the organization of chromatin within the nucleus and the regulation of individual genes has become an intensely studied subject . However , the complex nature of mammalian genomes has largely confounded efforts to understand the nature of this relationship . Several reports have catalogued the DNA and chromatin associated with the nuclear lamina and nucleolar periphery [15] , [16] , [25] . These findings have identified common characteristics of each domain , yet the basis for their presence at these compartments has remained less clear . Other studies have utilized fusion proteins to artificially tether lacO arrays to the nuclear lamina and other nuclear bodies [26] , [27] , [28] . Conversely , we have identified an endogenous sequence element , utilizing native nuclear machinery , that is capable of influencing subnuclear position . While transgenes with binding sites for the vertebrate insulator protein CTCF [29] have been shown to associate with nucleoli in a CTCF-dependent manner [30] , it is not known how frequently endogenous CTCF sites recapitulate this phenomenon . Our data demonstrate that 5S rDNA sequence can confer a positional bias in localization , and correlates with an attentuation of nearby pol II transcription ( summarized in Figure 6 ) . Importantly , the localization of 5S rDNA pseudogenes to the nucleolar periphery suggest this event is not limited to ectopic transgene integrations . Biased conservation of transcription factor binding sites within 5S pseudogenes implies a functional role in their endogenous contexts . We propose that the internal transcription factor sites of 5S rDNA represents a novel , cis- acting influence of nuclear position in mammals . This hypothesis is supported by the observed enrichment of 5S rDNA sequences in nucleolar-associated chromatin of human cells [15] , [16] . Recently , genome-wide maps of pol III and associated transcription factor binding in human cells have suggested structural roles reminiscent of what has been observed in yeast . These studies identified a number of “extra-TFIIIC” ( ETC ) loci , TFIIIC-bound regions not associated with a pol III complex or transcription unit [31] , [32] . However , unlike the ETC loci of yeast , which are associated with silencing of nearby pol II-driven promoters , human ETC loci are correlated with active pol II genes . In contrast , we observed high levels of the repressive H3K9me3 modification surrounding the 5S rDNA sequence . Thus the functional properties of ETC loci appear to be distinct from the repressive effect on pol II transcription that we have observed for 5S rDNA . Importantly , this demonstrates that presence of the TFIIIC complex alone is not sufficient to explain the effect on neighboring pol II transcription , suggesting additional or alternative factors . For example , TFIIIC recruitment to 5S rDNA first requires the binding of the TFIIIA , which specifically recognizes the A and C boxes . Alternatively , the strong nucleosome positioning properties of 5S rDNA may play a role in its localization and repressive effects on neighboring pol II transcription . Collectively , these observations suggest broad and diverse roles for pol III genes and derived sequences in the organization of chromatin within the mammalian nucleus . Because of their number , pol III promoters may exert a stronger influence on structural organization than pol II-directed gene activity . As pol III activity is coupled with differentiation and cellular metabolism , association of pol III and transcription factors with elements such as the 5S pseudgoenes we have described , may provide the basis for global organizational and structural changes within the nucleus in response to external stimuli [33] .
E14 ES cells were cultured under standard conditions . To generate stable lines , ES cells were transfected with 1 µg of linearized plasmid using lipofectamine ( Invitrogen ) and selected in the presence of G418 for 14 days . We verified stable neomycin resistance for most lines by culturing with G418 and noted no increased levels of cell death . To induce differentiation , 2×105 ES cells were plated on 60 mm2 dishes without LIF and in the presence of 0 . 1 µM retinoic acid ( Sigma ) then cultured for 8 days , with passaging to maintain low cell density . For immunofluoresence and DNA FISH , cells were plated at low density and grown on coverslips 18–24 hours . Coverslips were permeabilized with cytoskeletal ( CSK ) buffer ( 100 mM NaCl , 300 mM sucrose , 3 mM MgCl2 , and 10 mM PIPES pH 6 . 8 ) , then fixed in 4% paraformaldehyde ( PFA , Electron Microscopy Sciences ) for 10 minutes at room temperature , washed twice for 5 minutes in 1× PBS ( Cellgro ) , then stored in 75% ethanol at 4°C . Coverslips were re-hydrated with several washes of 1× PBS prior to DNA FISH experiments . To generate Tg5S , the 5S rDNA sequence ( a gift of B . Solner-Webb , Johns-Hopkins University ) was cloned into a vector that contains the neomycin resistance gene under the control of the HSV promoter , and the Thymidine kinase gene under the control of the mouse Pgk1 promoter ( a gift of D . Ciavatta , University of North Carolina ) . Tg0 was the vector without the 5S rDNA insert . RNA was isolated from cultured cells with Trizol reagent ( Invitrogen ) , DNAsed ( RQ1 DNAse , Promega ) and 500 ng of total RNA was used for each reaction . Samples were reverse transcribed using random-hexamer primers , with Superscript II Reverse Transcriptase ( Invitrogen ) . Primers are listed in Table S2 . Tk mRNA levels were first normalized to Gapdh levels . Real-time quantitative PCR was carried out on 25 ng of cDNA , in triplicate for each gene , on an ABI 3700 ( Applied Biosystems ) , using the Fast SYBR Green Master Mix ( Applied Biosystems ) . Data was analyzed in Microsoft Excel ( Microsoft ) , and is shown as the log2-transformation of RNA levels relative to copy number . Statistical significance was determined by two-tailed t-test . Transgene integration sites were determined using one of two approaches . The Tg5S#6 insertion site was identified using the TAIL PCR protocol , with degenerate primers as described [34] . The insertion site for Tg5S#9 was identified using inverse-PCR . Briefly , 1 µg of DNA was digested with XbaI , then ligated overnight with T4 ligase ( NEB ) at 2 ng/µl . DNA was concentrated by ethanol precipitation , and 50 ng of the ligation was used in a nested PCR reaction . PCR products were purified from an agarose gel using a QIAGEN gel extraction kit ( QIAGEN ) and quantified on a QUBIT flourometer ( QIAGEN ) . PCR products were directly sequenced and analyzed by BLAST searches to the reference assembly of the mouse genome . Each insertion was confirmed by PCR . Primers are listed in Table S2 . sequence . Each vector was linearized with XhoI ( NEB ) prior to lipofection . Coverslips were rehydrated in 1× PBS , before blocking in 10 mg/ml IgG-free BSA ( Jackson Immunochemical ) and 0 . 2% Tween-20 ( Fisher ) for 20–30 minutes at room temperature . Rabbit anti-Nucleolin ( Bethyl Laboritories , A300-711A ) was added at 1∶400 dilution into blocking buffer and incubated overnight at 4°C . Coverslips were then washed with 1× PBS , and incubated with biotinylated goat-anti-rabbit antibody , diluted in blocking buffer at 1∶400 , for 2–3 hours at room temperature . Following washes with 1× PBS , cells were post-fixed with 2% PFA for 3 minutes at room temperature , washed extensively with 1× PBS , then treated with 0 . 01 mg/ml pepsin ( Sigma ) diluted in pre-warmed 0 . 01 N HCl for 5 minutes , and then washed extensively with 1× PBS . Following a dehydration series in ethanol , DNA was denatured in 70% formamide ( Ameresco ) and 2× SSC ( Cellgro ) for 10–20 minutes at 85°C . After several washes with cold 2× SSC , cells were incubated with prehybridized DNA FISH probes ( see below ) overnight at 37°C . Coverslips were washed twice with 50% formamide and 2× SSC , twice with 2× SSC ( one wash had 100 ng/ml DAPI added ) , once with 1× SSC . To detect biotinylated secondary antibodies , coverslips were then washed once with 4× SSC for 5 minutes , incubated for 20–30 minutes in Streptavidin-647 ( Invitrogen ) in 2 mg/ml BSA and 4× SSC , followed by 5 minute washes of 4× SSC , 4× SSC with 0 . 5% Tween-20 ( Fisher ) , and 4× SSC . All washes and incubations for biotin detection were carried out at 37°C . In addition to the vector backbone , the following BAC and fosmid probes were used in this study: BACs: 5∶134 ( RP24-193L24 ) , 7∶30 ( RP23-151J21 ) , 8∶126 ( RP24-372G15 ) , 10∶27 ( RP24-213F23 ) , 11∶74 ( RP23174M12 ) , Mid1 ( RP24-229F18 ) ; and fosmids: 6∶112 ( G135P69622C7 ) , 8∶48 ( G135P60371F8 , G135P60172E7 ) , 19∶19 ( G135P64778C12 ) , PAR ( G135P601180H2 ) ( all clones were acquired from CHORI BPRC ) . BACs and fosmids were isolated by a standard alkaline lysis protocol . Approximately 25 ng of DNA was labeled with BioPrime DNA labeling kit ( Invitrogen ) , using FITC-conjugated dUTP ( Roche ) , Cy3- or Cy5-conjugated dCTP ( GE Healthcare ) , and stored in 70% ethanol at −20°C . To prepare FISH probes for hybridization , probes were precipitated with mouse Cot-1 DNA ( Invitrogen ) , yeast tRNA ( Invitrogen ) , and Salmon Sperm DNA ( Invitrogen ) . After washes with 75% and 100% ethanol , probes were air-dried and denatured for 10 minutes in 50–100 µl of 100% formamide at 85°C . An equal volume of 2× hybridization buffer ( 25% dextran sulfate/4× SSC ) was then added , and probes were pre-hybridized for 60 to 90 minutes at 37°C . Probes were stored at −20°C until use . IF-DNA FISH was carried out as described in Methods . For transgene-nucleolus association , cells were visualized on Leica DMLB fluorescent microscope ( Leica ) , captured on a Retiga 2000R Fast camera ( Qimaging ) , using QCapture software ( Qimaging ) , and merged with Adobe Photoshop ( Adobe ) . DNA FISH signals were considered ‘nucleolar associated’ if the FISH signals were in contact with , or within , the Nucleolin signal . For determining nucleolar association of 5S rDNA , pseudogenes , and genomic loci with transgene insertions , Z-stacks of each channel were taken on a Ziess AxioImager M2 microscope , deconvolved using the Axiovision software package ( Zeiss ) , then rendered in 3-dimensions using the ZEN Light Edition 2009 software ( Zeiss ) . Signals were considered ‘internal’ , if the center of the FISH signal was internal to the Nucleolin signal; ‘peripheral’ if the pixels of the FISH and Nucleolin signals were overlapping , but the center of the FISH signal was outside; and ‘not associated’ if there was visible distance between the DNA FISH signal and the outside of the Nucleolin-labeled nucleolus . Statistical significance was determined by chi-squared . Copy number was determined by quantitative PCR to determine the number of Neo gene copies relative to an endogenous locus ( the Ascl2 promoter ) , then normalized to a genomic DNA sample containing 1 copy of Neo for each diploid genome . Primers are listed in Table S2 . ES cells were trypsinized , counted , resuspended at 107 cells/ml and fixed with 1% formaldehyde . After quenching with 0 . 125 M glycine , cells were pelleted , washed once with cold 1×PBS , pelleted again and used for ChIP or frozen at −80°C . Protease inhibitors ( Sigma ) and PMSF ( Sigma ) were added to all steps until washing steps . For GTF3C5 and pseudogenes , ChIP was performed as described [35] . To measure histone modificiations or GTF3C5 association with transgenes , cell pellets were resuspended in solution L1 ( 50 mM HEPES-KOH , pH 7 . 5 , 140 mM NaCl , 1 mM EDTA , 10% glycerol , 0 . 5% NP-40 , 0 . 25% Triton X-100 ) at 107 cells/ml , mixed at 15 minutes and gently pelleted at 4°C . Cell pellet was resuspended in solution L2 ( 10 mM Tris-HCl , pH 8 . 0 , 200 mM NaCl , 1 mM EDTA , 0 . 5 mM EGTA ) at 107 cells/ml , mixed at 15 minutes and gently pelleted at 4°C . Cells were lysed in solution L3 ( 10 mM Tris-HCl , pH 8 . 0 , 100 mM NaCl , 1 mM EDTA , 0 . 5 mM EGTA , 0 . 1% Na-Deoxycholate , 0 . 5% N-lauroylsarcosine ) for 10 minutes at 4°C . Chromatin was sheared by sonication to generate fragments 2–600 bp . Before immunoprecipitation , 1/10th of each sample was removed as ‘input’ . 5 µg of Antibody ( rabbit anti-GTF3C5 , A301-242A , Bethyl Laboratories; rabbit anti-H3K4me2 , 07-030 Millipore; mouse anti-H3K9me2 , ab1220 , Abcam; rabbit anti-H3K9me3 , ab8898 , Abcam; or mouse anti-H3L27me3 , ab6002 , Abcam ) or normal rabbit sera ( Abcam ) was conjugated to protein A/G beads in 0 . 5%BSA/1×PBS overnight at 4°C on a nutating platform . Chromatin was incubated with bead-conjugated primary antibody overnight at 4°C with gentle mixing . For GTF3C5 ChIP , beads were then washed for 5 minutes at 4°C with gentle mixing , using the following solutions: Low Salt Buffer ( 0 . 1% SDS , 1% Triton X-100 , 2 mM EDTA , 20 mM Tris , 150 mM NaCl ) , twice; High Salt Buffer ( 0 . 1% SDS , 1% Triton X-100 , 2 mM EDTA , 20 mM Tris , 500 mM NaCl ) , once; LiCl buffer ( 1 mM EDTA , 10 mM Tris , 250 mM LiCl , 1% NP-40 , 1% Na-Deoxycholate ) , twice; and TE ( 10 mM Tris , 1 mM EDTA ) , twice . For histone modifications , beads were washed 4 times with RIPA buffer , and once with TE containing 50 mM NaCl . Chromatin was eluted from beads with 2 , 15-minute washes at 65°C using freshly prepared Elution Buffer ( 1% SDS/0 . 1 M NaHCO3 ) . To isolate DNA , 5 M NaCl was added to pooled eluates or input chromatin to a final concentration of 0 . 2 M , and incubated for at least 4 hours at 65°C , then treated with 30 µg of Proteinase K ( Roche ) for 2 hours at 55°C . After addition of 10 µg linear acrylamide as a carrier ( Ambion ) , DNA was extracted with 25∶24∶1 phenol∶choloform∶isoamyl alcohol ( Sigma ) , precipitated with 100% ethanol , and resuspended in nuclease-free ddH20 ( Promega ) . For psuedogenes , three replicates of quantitative PCR were carried out on an ABI 3700 ( Applied Biosystems ) , using the Fast SYBR Green Master Mix ( Applied Biosystems ) . For transgene and 5S rDNA enrichment , 2–5 replicates were performed on Bio-Rad CDX96 instrument , a using SsoFast EvaGreen Supermix ( Bio-Rad ) . PCR primers are listed in Table S2 . Data are displayed as enrichment of amplicon relative to a negative control region in each ChIP . Data was analyzed in Microsoft Excel ( Microsoft ) ; statistical significance was determined by two-tailed t-test . | Eukaryotic genomes are compartmentalized within nuclei such that physiological events , including transcription and DNA replication , can efficiently occur . The mechanisms that regulate this organization represent an exciting , and equally enigmatic , subject of research . In mammals , the identification of elements that influence these associations has been impeded by the complex nature of the genomes . Here , we report the identification and characterization of such an element . We demonstrate that the integration of a 5S rDNA gene , a 119 base pair noncoding RNA transcribed by RNA polymerase III , into a new genomic location can significantly influence the association of the host region with the nucleolus . This positioning has drastic , inhibitory effects on the transcription of a neighboring protein coding gene transcribed by RNA polymerase II , demonstrating a functional relationship between localization and gene expression . We also provide data that suggest this may be an endogenous phenomenon , through a class of repetitive sequences derived from 5S rDNA . Together , our data not only demonstrate a structural role for 5S rDNA but also suggest that nuclear organization of mammalian genomes may be strongly influenced by repetitive sequences . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"cell",
"biology",
"cell",
"biology",
"chromosome",
"biology",
"gene",
"expression",
"genetics",
"epigenetics",
"biology",
"molecular",
"genetics",
"genetics",
"and",
"genomics"
] | 2012 | Nucleolar Association and Transcriptional Inhibition through 5S rDNA in Mammals |
Dengue virus ( DENV ) transmission is spatially heterogeneous . Hence , to stratify dengue prevalence in space may be an efficacious strategy to target surveillance and control efforts in a cost-effective manner particularly in Venezuela where dengue is hyperendemic and public health resources are scarce . Here , we determine hot spots of dengue seroprevalence and the risk factors associated with these clusters using local spatial statistics and a regression modeling approach . From August 2010 to January 2011 , a community-based cross-sectional study of 2012 individuals in 840 households was performed in high incidence neighborhoods of a dengue hyperendemic city in Venezuela . Local spatial statistics conducted at household- and block-level identified clusters of recent dengue seroprevalence ( 39 hot spot households and 9 hot spot blocks ) in all neighborhoods . However , no clusters were found for past dengue seroprevalence . Clustering of infection was detected at a very small scale ( 20-110m ) suggesting a high disease focal aggregation . Factors associated with living in a hot spot household were occupation ( being a domestic worker/housewife ( P = 0 . 002 ) , lower socio-economic status ( living in a shack ( P<0 . 001 ) , sharing a household with <7 people ( P = 0 . 004 ) , promoting potential vector breeding sites ( storing water in containers ( P = 0 . 024 ) , having litter outdoors ( P = 0 . 002 ) and mosquito preventive measures ( such as using repellent , P = 0 . 011 ) . Similarly , low socio-economic status ( living in crowded conditions , P<0 . 001 ) , having an occupation of domestic worker/housewife ( P = 0 . 012 ) and not using certain preventive measures against mosquitoes ( P<0 . 05 ) were directly associated with living in a hot spot block . Our findings contribute to a better comprehension of the spatial dynamics of dengue by assessing the relationship between disease clusters and their risk factors . These results can inform health authorities in the design of surveillance and control activities . Focalizing dengue control measures during epidemic and inter-epidemic periods to disease high risk zones at household and neighborhood-level may significantly reduce virus transmission in comparison to random interventions .
The incidence of dengue , a vector-borne viral disease , has risen markedly in the last decades affecting more than half of the world’s population [1] . According to a recent study , 390 million dengue infections are estimated to occur annually [2] . Dengue and its vectors have spread into previously unaffected areas and presently this disease is endemic in 128 countries [3 , 4] . Currently , dengue control methods rely mostly on vector reduction; however , these activities have proven largely unsuccessful [4] . Geographic information systems ( GIS ) and spatial analysis techniques are important tools for public health as they integrate the detection of disease spatial patterns , the identification of unusual aggregations ( hot spots ) of epidemiological events and allow the prediction of high risk areas of disease transmission [5 , 6] . Dengue hot spots identification is suitable to focalize health control measures and epidemiological surveillance in a cost effective manner particularly in regions where resources are limited [7 , 8] . Dengue virus ( DENV ) belongs to the Flavivirus genus of the family Flaviviridae [9] . It is transmitted by the bite of infected female Aedes mosquitoes , mainly Ae . aegypti [10] . Although Ae . albopictus is a less efficient vector; it has also been related to dengue outbreaks [11] . DENV consists of four serologically distinct serotypes ( DENV-1 to -4 ) each of them capable of causing the entire range of dengue-related disease symptoms [12] . In Venezuela , dengue has become a major public health problem of urban areas . DENV transmission is endemic with the co-circulation of the 4 viral serotypes [13] . Control of this infection and of its mosquito vector has proven challenging due to growing population density , increasingly crowded living conditions , unreliable water supply , and enduring problems in public services [14 , 15 , 16] . Furthermore , in recent years Venezuela experienced an increase in dengue incidence; with this increase being related to health sector crisis , budget cuts and shortage of medicines due to technical and economical limitations [16 , 17 , 18] . Despite control measures , transmission of dengue in Venezuela has become persistent with an average of 40 , 000 cases annually in non-epidemic years and three large epidemics in the past decade [19] . The most recent and biggest occurred in 2010 , where approximately 125 , 000 cases including 10 , 300 ( 8 . 6% ) with severe manifestations were registered [19] . Maracay city in Aragua state has become one of the most important endemic urban areas in the country . The highest number of cases and dengue incidence during the 2010 epidemic was reported in Aragua ( 495 cases per 100 . 000 inhabitants ) [14 , 19] . Indeed , during this year , Venezuela was the country with the third highest number of reported dengue cases in the Americas and ranked second in the number of severe cases [20] . National dengue control measures involve strategies to reduce the vector [21 , 22] however , in recent years surveillance and control measures have been applied irregularly or have been absent [17 , 18] . Previous studies in Maracay , a dengue hyperendemic city in Venezuela , have shown that certain areas are more prone to maintain higher dengue transmission and for longer periods than others [15] indicating that some epidemiological conditions are stable through time . Using mapping technology and spatial analysis of epidemiological and seroprevalence data we attempt to draw risk-maps at a fine scale to identify clusters ( hot spots ) of DENV transmission within high dengue incidence neighborhoods in Maracay and relate them with the risk factors present in the studied areas . Results will be used to inform health authorities to improve dengue control strategies .
Maracay is the capital city of Aragua state in the north-central region of Venezuela ( 10° 15′ 6″ N , 67° 36′ 5″ W ) with an estimated population of 1 . 139 . 000 inhabitants [23] . The annual average temperature is 25 . 5°C ( min 19°C , max 31°C ) with 74% of humidity and an annual precipitation of 910 mm [24] with two seasons , a dry ( November-April ) and a rainy season ( May-October ) . Three neighborhoods within two municipalities of high dengue incidence [15] were chosen for our study ( Fig 1 ) . The reported dengue incidence is slightly higher in Mario Briceño Iragorry municipality than in Girardot municipality ( Fig 2 ) . Lately , two dengue transmission peaks took place in these municipalities; one in 2007 and the second in 2009–2010 parallel to the whole dengue incidence in Aragua State and Venezuela as shown previously [14] . Caña de Azúcar and Candelaria neighborhoods , belonging to Mario Briceño Iragorry municipality , are located close to each other in the north-western area of Maracay . This municipality has a population of 99 , 852 inhabitants in an area of 54 km2 [23] . Caña de Azúcar and Candelaria neighborhoods have an area of 0 , 50 and 0 , 87 km2 , respectively; are divided by the “Limón” river and surrounded northerly by the mountainous National park “Henri Pittier” ( Fig 1 ) . Cooperativa neighborhood ( Girardot municipality ) is found in the north-east side of Maracay with an area of 1 . 1 km2 and the river “Las Delicias” running along its eastern border ( Fig 1 ) [25] . Girardot municipality has a population of 590 , 679 inhabitants in an area of 312 km2 [23] . Both municipalities are located within the metropolitan area of Maracay which comprises pre-planned urban areas . However , unplanned developments are also present and are characterized by the lack of public services such as electricity , water supply and garbage collection . In addition , piped-water supply is irregular in the entire Aragua state compelling the population to store water in tanks and other containers in order to ensure constant access to water [14 , 15] . A baseline cross-sectional study was carried out during the recruitment process of a prospective community-based cohort study described elsewhere [14] . Briefly , 2012 individuals aged 5–30 years inhabiting 840 households within the above mentioned neighborhoods were enrolled , from August 2010 to January 2011 , through house-to-house visits . The inclusion criteria were described previously [14] , briefly: age between 5–30 years old; living in the study area with no intention to move in the following 3 years; and consenting to attend the designated health centre in case of any symptoms . The scope of the study was clearly explained to all members of the household and the individuals were enrolled after written informed consent . Data were collected through an individual and a household structured questionnaires . The individual questionnaire contained data related to socio-demographic , epidemiologic and clinical history while socio-economic and environmental factors were recorded with the household questionnaire . Serological and hematological data were acquired through blood sample collections [14] . The geographical location of each household was obtained using a hand-held Global Positioning System ( GPS , Garmin Ltd . ) . As reported earlier [14] , 5–10% of households with a probable higher socio-economic status in Cooperativa refused to participate . Since socio-economic variables were similar between Cooperativa and Candelaria where refusal was minimal , we believe that selection bias in Cooperativa was small . In the present study , 1985 individuals living in 837 households that had a recorded geographical position were included . They were distributed as follows: Candelaria with 452 individuals living in 208 households , Cooperativa with 601 people in 266 households and Caña de Azucar where 932 subjects inhabited 363 residences . A 10 ml blood sample was collected from each enrolled individual to perform baseline dengue serology and a full blood count . Dengue seropositivity was determined using the Hemaglutination Inhibition ( HI ) Assay as described in detail in Velasco et al . , [14] . Two variables for previous dengue infection were defined: a ) Past dengue infection: HI titres >1:20 , and b ) Recent dengue infection: HI titres ≥ 1280 [14 , 26] . As reported earlier , 77 . 4% of the population under study had a past dengue infection while 10% exhibited a recent infection . The latter was more prevalent in Caña de Azucar ( 12 . 8% ) , followed by Cooperativa ( 8 . 1% ) and Candelaria ( 7 . 4% ) neighborhoods [14] . The hypothesis that an event of dengue infection is equally likely to occur at any location within the study area , regardless of the locations of other events , was tested . For that , we used one local measure of spatial association , the local Getis statistic [27] . The event was analyzed at two spatial scales: 1 ) at household level , and 2 ) at block level . Dengue seroprevalence at block and household level was standardized as the total number of seropositive individuals divided by the total number of individuals surveyed in a block or household , respectively . Risk maps at block and household level were drawn for the two seroprevalence outcome variables . The local Getis statistic , Gi* ( d ) detected significant local clustering of high positive ( hot spots ) values of dengue prevalence around each point ( e . g . household infection ) within a radius ( circular window ) of specified distance d from that location . The distance d defined the neighborhood search for a particular house or block , with nearby locations being expected to have similar values . The value obtained was compared ( by using the Monte Carlo randomization procedure ) with the statistic’s expected value to indicate if the degree of clustering of dengue prevalence in the vicinity of a particular location was greater or less than expected by chance . To correct for multiple comparisons when using Gi* ( d ) , significance levels ( P < 0 . 05 ) were adjusted according to Ord & Getis ( 1995 ) [28] . We calculated Gi* ( d ) at different window sizes with the maximum Gi* ( d ) distance corresponding to the scale at which the Gi* ( d ) maximum value was found that is , the scale of the spatial dependence of the process under study [29] . The analyses of Gi* ( d ) were carried out through the Point Pattern Analysis ( PPA 1 . 0 , San Diego State University , San Diego , CA http://www . acsu . buffalo . edu/~geojared/tools . htm ) . The results were shown in maps using the softwares QGIS 1 . 8 . 0-Lisboa ( GNU—General Public License ) and ArcGIS 10 ( ESRI Corporation , Redlands , CA ) . The satellite images of each neighborhood were obtained from Google EarthTM . Two outcome variables were defined based on the Getis analysis of recent dengue seroprevalence hot spot detection: 1 ) Individuals living in a hot spot household , and 2 ) Individuals living in a hot spot block . Univariate and multivariate analyses of potential risk factors associated to hot spots at household and block level of recent dengue infection were performed using SPSS ( SPSS Inc . , version 20 . 0 , Chicago , Illinois ) software . Variables included in the analysis were the following: demographic ( age , gender , occupation ) , socio-economic ( duration of residence , studying and having a job , type of housing , number of household rooms , persons per household ) , crowding ( number of persons living in a household divided by the number of household rooms ) , environmental ( water storage at home , availability of public services , presence of litter , used car tires and bottles outdoor and indoor flower vases ) , and mosquito preventive measures ( screened windows/doors , use of mosquito nets , insecticide and repellent usage , and container washing ) . The studied variables were previously described in detail [14] . Variables considered as confounders were age and gender . Continuous variables were converted into ordered categorical variables when suitable; otherwise they were dichotomized above and below their mean value ( if normally distributed ) or the median ( when non-normally distributed ) . The variable crowding was divided into quartiles and a cut-off point was set between the third and fourth quartiles ( ~1 . 5persons/room ) where a difference in prevalence was observed . Proportions were compared using chi-square tests . Fisher’s exact test was used when one or more cells of the contingency table had an expected count of less than five . Logistic regression was used to compare crude and adjusted odds ratios ( OR ) . Significance was determined at the 5% level ( P < 0 . 05 ) . The final models contained variables independently associated with living in a hot spot household or hot spot block . Data were analyzed anonymously and individuals were coded with unique numeric identifiers . All adult subjects ≥18 years old provided written informed consent , and a parent or guardian of any child participant provided written informed consent on their behalf . Children between 8 and 17 years old provided written informed assent [14] . The study was approved by the Ethics Review Committee of the Biomedical Research Institute , Carabobo University ( Aval Bioetico #CBIIB ( UC ) -014 ) , Maracay , Venezuela; the Ethics , Bioethics and Biodiversity Committee ( CEBioBio ) of the National Foundation for Science , Technology and Innovation ( FONACIT ) of the Ministry of Science , Technology and Innovation , Caracas , Venezuela; and by the Regional Health authorities of Aragua State ( CORPOSALUD Aragua ) . The study was conducted according to the principles expressed in the Declaration of Helsinki [30] .
Risk maps of the spatial distribution and local clustering of dengue seroprevalence for each outcome variable ( past and recent ) at block and household level are shown in Figs 3 and 4 . A high seroprevalence of past dengue infection was found across all neighborhoods , resulting in risk maps where most blocks exhibited a seroprevalence >40% . Consequently , no past dengue seroprevalence clusters were detected by the local spatial statistics at any spatial scale ( household or block; Figs 3a and 4a ) . Risk maps of recent dengue seroprevalence depicted a greater spatial heterogeneity . The highest frequency of recent dengue transmission at block level ( seroprevalence >36% ) was located in the southern part of Caña de Azucar neighborhood , while the majority of Candelaria’s blocks and the rest of Caña de Azucar displayed a seroprevalence below 21% ( Fig 3b ) . Most blocks within La Cooperativa neighborhood showed low seroprevalence except for three areas , one in the northern , the second one in the middle and the third one in the southern edge where a prevalence higher than 36% was found ( Fig 4b ) . Significant hot spots at household and block level were identified for recent dengue seroprevalence . Spatial statistics showed that most of the recently infected individuals were spatially located toward the southern side of Caña de Azucar neighborhood ( Fig 3b ) . The local Getis statistic identified in this southern area 3 hot spots at block level ( blocks 86 , 88 and 92 ) and 9 hot spots at household level ( see below and Fig 3b ) . Consequently , the most relevant recent dengue transmission gathering of clusters was found in this particular neighborhood . Additionally , hot spots at household and block level were detected by the local Getis analysis in all neighborhoods with a total of 65 individuals residing in the 39 detected hot spot households while 144 people lived in 62 households within 9 hot spot blocks . Four clusters at block level were identified in Caña de Azúcar , one in Candelaria and four in La Cooperativa neighbourhood ( Figs 3b and 4b ) . The four hot spots at block level found in Caña de Azucar contained 36 houses , 13 ( 36% ) of which were also hot spots households . The cluster found in Candelaria contained 9 households and one ( 11% ) hot spot household . In La Cooperativa neighborhood , four hot spots blocks were identified including 17 households of which 4 ( 23 . 5% ) were hot spots . Finally , the spatial-scale ( maximum Gi* ( d ) distance ) at which all dengue clusters were detected varied between 20 meters for household level clusters and 90–110 meters for block level clusters suggesting the relevant spatial scale at which dengue transmission occurs in the studied urban landscape . The values of the two outcome variables of recent dengue seroprevalence hot spot detection resulted as follows: Sixty five individuals living in hot spot households ( n = 65/1985 = 3 . 3% ) , and 144 individuals living in hot spot blocks ( n = 144/1985 = 7 . 3% ) .
To our knowledge , this is the first study to apply spatial analysis techniques in Venezuela to determine areas of higher transmission of dengue coupled with the identification of risk factors that may explain the higher endemicity within these clusters . These methods are increasingly being used to understand dengue epidemiology with few studies published so far in this area of research in the Americas [31 , 32 , 33 , 34 , 35] . We detected spatial clusters of dengue seroprevalence and identified the risk factors for dengue transmission associated with these clusters in a dengue hyperendemic city in Venezuela . Hot spots of recent dengue infection at household and block level occurred in all 3 neighborhoods under study . However , one neighborhood , Caña de Azucar , concentrated the majority of hot spots and accounted for most dengue transmission across the whole studied area showing the focal nature of this mosquito-borne viral infection . Indeed , our results indicated that most of the clustering distance did not extend beyond 100 m suggesting that an underlying spatial process of dengue transmission is acting at such small scale . Conditions that enhanced the risk of transmission and infection by dengue virus in a hot spot household or a hot spot block were related with occupation ( being a domestic worker/housewife ) , lower socio-economic status ( to live in a shack , crowded conditions , more people per room ) , the creation of potential mosquito breeding sites ( to store water in containers and having litter outdoors ) and mosquito preventive measures ( have screened windows/doors , the usage of insecticide or repellent , and container washing ) . Spatial analysis techniques applied to vector borne diseases have proven useful to define high risk areas of transmission and factors associated with this risk , while informing health authorities on better targeted control actions as well as generating models applicable to other regions [15 , 36 , 37 , 38] . Using seroprevalence data from three neighborhoods in Maracay city [14] we drew risk maps and determined hot spots of dengue transmission at household and block level . An important proportion of dengue infections are clinically inapparent [39 , 40] contributing to increased viral transmission . Therefore , the use of seroprevalence data over incidence ( symptomatic cases ) may give a more accurate estimate of high dengue transmission areas . Risk maps of past dengue seroprevalence showed a certain spatial homogeneity as the majority of blocks had a past seroprevalence > 40% following the high overall prevalence found in the population under study ( 77 . 4% ) [14] ( Figs 3a and 4a ) . This also resulted in the absence of identifiable hot spots for past dengue at any spatial level . The neighborhoods included in our study have been reported as areas of perennial dengue transmission , tending to maintain an infectious cycle of dengue outside of the rainy season [15 , 41] . The progressive entry of the different serotypes of dengue in Venezuela has been associated with the major epidemics in the country [14 , 41 , 42 , 43 , 44] . In 2010 , our fieldwork coincided with one of the major epidemics of dengue in Venezuela [19] probably resulting in the observed high seroprevalence . Similar observations of dengue post-epidemic prevalence were observed in American Samoa in 2010 with a seroprevalence of 95 . 6% [45] . Spatial heterogeneity was revealed when drawing risk maps of recent dengue seroprevalence . Recent dengue infections were recorded in 10% ( n = 200 ) of the individuals under study [14] . Interestingly , areas with high recent seroprevalence coincided with areas of increased past dengue seroprevalence which could indicate the persistence of dengue transmission in these locations . A total of 39 significant hot spots for recent dengue at household level and 9 significant hot spots at block level were found , most of them in Caña de Azucar neighborhood , which confirms the higher risk of recent dengue transmission in this neighborhood ( Fig 3b ) . Caña de Azúcar neighborhood is one of the most densely populated areas of Maracay and living in this neighborhood was associated to a lower socio-economic status and a higher proportion of potential breeding sites [14] . Individuals living in this neighborhood were significantly more prone to store water at home , live in smaller houses and in more crowded conditions than people residing in the other two neighborhoods [14] , This findings suggest an increased chance of dengue transmission in this area [12 , 31] . The transmission of dengue in our area of study was highly focal ( radius = 20–110 meters ) suggesting that at domestic level the necessary conditions for oviposition , growth , feeding and reproduction of the mosquito vector exist . In agreement with our findings , other studies have reported a short flight range for Ae . aegypti ( radius <40 m ) where the vector tends to be spatially clustered at household level in relation to the occurrence of indoor breeding sites [46] and abundant human hosts [47] . Other authors found that female mosquitoes do not visit more than 3 houses in a lifetime [48] . In Iquitos , Peru , researchers registered a mosquito flight range of 10–30 meters , and concluded that in this area the vector does not fly away from the water containers where they breed [49] . Likewise , it has been reported that dengue cases cluster within households [50] . The small spatial dependency scale ( 20–110 meters ) and the finding that people who spend more time indoors such as domestic workers/housewives were at a higher risk of recent dengue ( Tables 5 & 6 ) compared to those who had jobs away from home strongly indicates that transmission occurs mainly at home as suggested in other studies [14 , 51 , 52] . Here , it is important to point out that our study is focused mainly on the local spatial scale which is much related to the short-range flight of mosquito dispersal; however , we also highlight the relevance of the human movement to the spatial dynamic of dengue at large spatial scale . It is the interaction between infected mosquito dispersal at very small scale and infected human movements at large scale that underlie the dynamics of dengue transmission through space and time . Proxy markers of poverty or lower socio-economic status were strongly associated with hot spot households and blocks and were more relevant spatially than those found in a previous study with classical analysis of seroprevalence data [14] . People living in a shack were associated with living in a hot spot household , while those in smaller households ( <5 rooms ) and in crowded conditions were related to inhabiting a hot spot block ( Tables 5 & 6 ) . In our studied neighborhoods , population growth and the need for extra income resulted in the sub-division of one-family dwellings into smaller “apartments” to house two or more families . The consequences are more crammed living conditions and deprivation [53] . Studies in Brazil [32] and Ecuador [54] determined that dengue clusters or hot spots are mainly located in poor areas . Unplanned urbanization and precarious living circumstances are characterized by the lack of proper public services ( piped water supply , electricity , garbage collection , sewage ) and crowding favoring the transmission of dengue [14 , 34 , 55] . Other authors found that increased dengue transmission was associated with socio-economic factors rather than climatic factors [56] . We explored the mosquito preventive measures taken by the population under study in order to understand if they had any effect in reducing dengue transmission . While some measures showed a protective effect , i . e . , individuals that used insecticide sprays and regularly washed their water containers were less prone to inhabit a hot spot block , other actions showed an opposite result . People that lived in households with screened windows/doors and that used mosquito repellent had a higher probability of residing in a hot spot block or household , respectively . Although this may be difficult to explain , we hypothesize that the latter results may be indirect markers of higher mosquito density prompting individuals to implement these protective measures , as reported by other authors [57] . Potential breeding sites were only found independently associated with hot spots at household but not at block level in multivariate analysis . The presence of litter in household premises and storing water in containers enhanced the odds of living in a hot spot household at least twice ( Table 5 ) . Several studies have linked the presence of cans , small canisters or other types of containers accumulated in and around household grounds as potential breeding sites that may be extending the possibilities of oviposition of dengue vector mosquitoes beyond the rainy season [36 , 58 , 59] . Poor piped water supply involves storing water in diverse type of containers creating a suitable environment for the growth and development of Aedes spp . especially during the dry season , consequently keeping a perennial dengue transmission in the population [15 , 47 , 60 , 61] . Storing water in containers in Venezuela is a practice that has been reported for several years [36 , 60] . The neighborhoods of our study have been subjected to long-lasting deficits in public services , especially prolonged interruptions of piped water supply and electricity . This has prompted the population to maintain water storage at home all year round . The worsening socio-economic situation in Venezuela can predict the perennial maintenance of dengue transmission and an increase in the frequency of epidemics as seen recently [19] . Targeting the identified hot spot areas with strategies such as source reduction and community education [49] may result in a cost-effective manner to improve dengue control in Venezuela and similar endemic areas . A limitation of our work is the absence of entomological data that may complement our results and provide more clues about the focal transmission of dengue in these urban areas . However , our results can be supported by the identified risk factors for dengue infection performed in the same population in a previous study [14] .
The application of geographic information systems and spatial analysis for the detection of areas of greater transmission of DENV is of vital importance for the prevention and control of vector-borne diseases such as dengue . We used seroprevalence data to understand the spatial spread and clustering of dengue in a hyperendemic city in Venezuela . In contrast with incidence data , seroprevalence has the advantage to include inapparent as well as symptomatic infections giving a more realistic view of the high transmission risk areas . Moreover , we performed an analysis of risk factors at a fine spatial scale comparing individuals living within hot spots at household and block level versus those that lived outside these geographical areas . We determined that most hot spots clustered in one neighborhood and that transmission occurs at a very small scale ( radius of 20–110 m ) and domestically giving the possibility to target the scarce resources to this specific area . Secondly , poverty-related factors and those related to potential breeding sites were associated with these hot spots pointing to measures that can and should be taken promptly as reported previously [15] . However , political will and channeling enough resources to alleviate low socio-economic conditions and improve public services are essential for success in dengue control . The identification of hot spots of dengue transmission and the factors associated with these clusters are important tools to inform health authorities to improve and target control measures against dengue . Further studies are needed to define if these hot spot areas are maintained through time while similar studies can be applied to other vector-borne infections such as malaria , chikungunya and Zika virus affecting both Venezuela and other Latin American countries . | Dengue is a mosquito-borne viral disease of global impact . In Venezuela , dengue is endemic with the co-circulation of the 4 viral serotypes and has become one of the most important public health problems of urban areas . During 2010 and 2011 , a baseline cross-sectional study was carried out as part of a dengue prospective cohort study . We enrolled 2012 individuals aged 5–30 years living in 840 households within 3 neighbourhoods in Maracay , one of the cities with the highest number of reported dengue cases in Venezuela . Serological data were obtained through blood sample collections prior to informed consent . Analyses of risk-maps at a fine scale were carried out to detect dengue seroprevalence hot spots ( areas of greater transmission ) within these neighborhoods and to relate them with potential transmission risk factors . Recent dengue infection clustered within and around households and blocks ( radius 20-110m ) , suggesting the relevant spatial scale at which disease transmission occurs in the studied area . People that lived within these hot spots were poorer and had more man-made potential mosquito breeding sites in and around their premises . Focalizing dengue control measures to infection high risk zones may result in a more cost effective approach of dengue surveillance and control . Spatial statistics analyses are powerful tools to identify the past and the actual distribution of dengue , localize high risk areas and help focalizing control measures . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusions"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"water",
"resources",
"venezuela",
"geographical",
"locations",
"vector-borne",
"diseases",
"animals",
"geoinformatics",
"infectious",
"disease",
"control",
"insect",
... | 2017 | Spatial Analysis of Dengue Seroprevalence and Modeling of Transmission Risk Factors in a Dengue Hyperendemic City of Venezuela |
A major challenge of eco-epidemiology is to determine which factors promote the transmission of infectious diseases and to establish risk maps that can be used by public health authorities . The geographic predictions resulting from ecological niche modelling have been widely used for modelling the future dispersion of vectors based on the occurrence records and the potential prevalence of the disease . The establishment of risk maps for disease systems with complex cycles such as cutaneous leishmaniasis ( CL ) can be very challenging due to the many inference networks between large sets of host and vector species , with considerable heterogeneity in disease patterns in space and time . One novelty in the present study is the use of human CL cases to predict the risk of leishmaniasis occurrence in response to anthropogenic , climatic and environmental factors at two different scales , in the Neotropical moist forest biome ( Amazonian basin and surrounding forest ecosystems ) and in the surrounding region of French Guiana . With a consistent data set never used before and a conceptual and methodological framework for interpreting data cases , we obtained risk maps with high statistical support . The predominantly identified human CL risk areas are those where the human impact on the environment is significant , associated with less contributory climatic and ecological factors . For both models this study highlights the importance of considering the anthropogenic drivers for disease risk assessment in human , although CL is mainly linked to the sylvatic and peri-urban cycle in Meso and South America .
Vector-borne diseases that threaten one-third of the world's population are driven by intertwined socio-economic and environmental factors , such as climate change and modifications of ecosystems through deforestation , conversion of natural habitats to man-made ecosystems and extended urbanisation [1] . To understand these disease agent dynamics , it is necessary to determine ( 1 ) the geographic area and associated ecological conditions where the transmission cycle could likely occur , with the infected vectors and host reservoirs , ( 2 ) the risk factors that promote transmission to humans and ( 3 ) the human communities that are the most exposed to infection hazards on a local scale [1–3] . Landscape ecology may contribute to the knowledge of the influence of biotic and abiotic factors on the presence and dynamics of the vectors and host reservoirs [4] . It also favours the development of spatial models of risk prediction at a relevant geographic scale [5] , which finds its theoretical and more practical extensions within the new pathogeography paradigm [6] . These spatial models theoretically make it possible to reveal the geographical areas where the transmission rate of the disease risk is predicted to be the highest by identifying the environmental , climatic and socio-economic risk factors that may expose the most vulnerable individuals and populations to microbial hazards and threats [7 , 8] . These models may summarise the concept of risk in epidemiology underlying the notions of hazard , exposure and vulnerability . Hazard represents at least the occurrence and distribution of the microbial agent under scrutiny in a geographical area as well as the distribution of vectors , hosts and their interaction . Exposure is related to the probability of an individual or a community being exposed to microbial hazard through recreational or occupational activities . Vulnerability represents the individual and group conditions that make humans more sensitive to infection , e . g . , genetic susceptibility or malnourished people [9] . Within the last decade or so , ecological niche models ( ENMs ) have been proposed in landscape epidemiology to explore the relationships between the potential distribution of vectors or host species reservoirs and environmental variables [10] . The ENMs are used to circumvent gaps in knowledge of species distribution and are based on the occurrence of a species and relevant environmental variables for identifying the most favourable habitats for the establishment and survival of the species of interest [11] . Then they project the relationships over a geographical area to identify non-surveyed areas where there are favourable environmental conditions , and which are propitious for the development and spread of this species . Applied to hosts [12] and vectors [7] of pathogens , it has been possible to better understand the complex influences of spatial heterogeneity and environmental variation on the distributions of species involved in the disease agent transmission cycle , often interpreted as the more likely distribution of the disease agent and hence the disease [13] . Within this framework , the vector-borne disease models show that at larger scales , vectors presence is correlated with climatic and non-climatic factors , with these abiotic factors having a strong influence on vector species range delineation , i . e . , the limits of distributional ranges towards more northern areas [7 , 14] . The influence of anthropic pressures on the environment plays a significant role at more local geographic scales and can unbalance the complex interactions between hosts , vectors and disease agents [15 , 16] . To properly identify the set of biotic and abiotic conditions suitable for disease maintenance and dispersal , the BAM ( biotic , abiotic , movement ) framework was proposed [17] . Biotic and abiotic conditions are based on transmission pathways between host and vector communities and shape the geographic and ecologic distributions of the parasite . The movement summarise limitations , accessibility and possible barriers for spreading opportunities . As such , ENMs applied to vector or reservoir-borne infectious diseases may be confounded to the hazards component part in disease risk calculation . This theoretical framework may help to choose the candidate biotic and abiotic variables and the scales at which all these components must be tested to best fit with the biological model . However , relevant movement may be complicated to model . Today , the development of risk maps for ( zoonotic ) vector-borne diseases remains difficult for two reasons . First , creating a risk map requires considering the notions of hazard , exposure and vulnerability , in addition to choosing the explanatory variables using the BAM framework . Indeed , the likelihood of contact and contamination between human and host-vectors can vary considerably from one region to another , depending on biodiversity and landscape management programs , education level , health surveillance and control , living conditions , economic resources , etc . [16] . Some anthropogenic variables such as the human footprint ( HFP ) , deforestation , urban expansion and poverty [18] allow studying the vulnerability of human communities . Second , for disease systems with multi-host species and/or multi-vector species [19] it may be unrealistic to model all the actors in systems of such diversified communities of vectors and hosts [20 , 21] . Identifying explanatory variables and modelling the occurrence of recognised vectors and/or hosts may miss important parts of the infectious disease system , leading to conflicting issues when suitable areas for disease agent establishment are expected to be considered as epidemiologic risks [22–24] . An alternative approach may be to focus on the occurrence of human cases , considering that disease records indicate the circulation of the pathogen , whatever hosts and vectors , including secondary ones , are involved in the disease agent’s life cycle [5 , 6] . In disease ecology , in the past decade these models relying on human case have shown relevance in identifying more favourable areas for diseases occurrence and risk prediction [25 , 26] . Thus , species distribution modelling ( SDM ) with human cases and climatic , environmental and anthropogenic variables may be useful in identifying the different factors influencing the complex disease transmission cycle such as for cutaneous leishmaniasis ( CL ) . CL is caused by a protozoan parasite of the genus Leishmania with a complex life cycle involving multiple phlebotomines and mammal species acting as natural vectors and reservoirs , respectively , for the parasite [27 , 28] . In Meso and Southern Americas , 940 , 396 new cases of cutaneous ( CL ) and mucosal leishmaniasis were reported by 17 endemic countries from 2001 to 2017 [29] . American cutaneous leishmaniasis is widespread in the Amazonian Basin and throughout the Neotropical rainforest biome , a region with high biodiversity , and caused by several Leishmaniinae species [30–35] . Within Amazonia , the different Leishmania species have a more focal distribution due to their transmission cycles associated with specific ranges of the host reservoirs and vectors [2] . Further , transmission cycles are mainly sylvatic , although urbanisation processes have been reported in some South American countries such as Colombia [34 , 36] . The sylvatic cycle occurs in forested environments and the rural/domestic cycle occurs mainly in forested-associated human settlements by intra-domiciliary transmission . At the infection focus ( a given area where transmission occurred ) , all components of the cycle must be brought together . Risk models aim to correlate these infection foci with human activities to define the areas that are at high eco-epidemiological risk of infection for humans . However , for leishmaniasis Vélez et al . ( 2017 ) [2] pointed out that the limit of these infection foci was complex to define due to ( 1 ) the high diversity of phlebotomine species and the numerous host species involved in the disease life cycle , ( 2 ) the diversity of Leishmania species , ( 3 ) the complexity of confirming phlebotomine species as vectors and wild mammalians as hosts and ( 4 ) the challenge of diagnosing human cases with clinical forms of leishmaniasis . Further , the large geographic extent of the disease and disease agent cycles that may operate in space induce many complex ecological interactions [36] and add uncertainty on the place of infections , which is problematic when models are based on the geolocation of human cases . Last , major anthropogenic disturbances in the Amazonian region impact complex networks of species communities in forest ecosystems; land uses and modifications of the natural habitats are recognised critical factors affecting the mammals and the phlebotomine community's abundance and density [37] . Previous studies have used SDM to map CL occurrence with human cases as input data based on the boosted regression tree ( BRT ) [14 , 38] and regression Bayesian modelling [39] showed that climatic parameters acted as the most important predictors of CL distribution at the scale of the South American continent [14 , 38] and in Brazil only [39] . However , beyond the climatic influence , the level of anthropogenic pressure can act at a finer local scale to influence the disease distribution cycle [40 , 41] . The aim of the present study was to map the risk of CL based at two different scales in the Amazonian forest and surrounding Neotropical moist forest ecosystems . This geographic area allows working at higher spatial resolution than previously published studies , controlling the influence of bioclimatic factors previously identified as disease occurrence drivers [14 , 38 , 39] and likely highlighting a putative role of more local bioecological drivers . We used maximum entropy implemented with the MaxEnt software [42] , based on a presence-background ENM , identifying non-linear responses of CL cases to different fine-resolution biotic and abiotic variables at both the Amazonian and French Guianan scales . These two models were run independently and are not assumed to validate each other , but instead are expected to show the extent to which the geographic grain influences the relative importance of contributory variables for the spatial prediction of the disease risk . We used only the official human CL epidemiological records as input data to predict the risk of leishmaniasis occurrence . The cases were geolocated in the health centres , resulting in uncertainty as to the contamination area and geography-biased case reports for this sylvatic disease . To stay within the BAM reasoning framework , we attempted to adapt the model to the real ecological conditions of the CL cycle . To reflect the most likely places of contamination and properly handle the field realities , we randomly distributed the occurrence of cases outside urban centres . By eliminating areas where one is unlikely to find autochthonous CL cases , we succeeded in integrating the movement ( M ) of the BAM framework . Several redistribution methods made it possible to control the sampling biases related to the uncertainty of case geolocation . The novelty of this work was its redistribution of the occurrences of the disease cases , testing several CL case distribution methodologies , to approach the ecological characteristics of the disease as closely as possible .
For the Amazonian model , we used a total of 149 , 368 human CL cases referenced in 1415 localities from Brazil , Colombia and French Guiana . These case records were predominantly located in the same large Neotropical moist forest biome that encompasses the Amazonian basin , the Guiana shield , and north-west forests of South America ( Fig 1 ) . For Brazil , 75 , 441 CL cases , reported from 2007 to 2015 , spread across 444 localities in the Amazonian states of Acre , Rondônia , Tocantins , Pará , Roraima , Amapá , Mato Grosso and Amazonas were obtained from the Secretaria de Vigilância em Saúde-SVS ( Secretary of Surveillance in Health ) from the Brazilian Ministry of Health . The data were validated by the Technical Group of Leishmaniasis , the Coordenação Geral de Doenças Transmissíveis ( CGDT ) , the Departamento de Vigilância de Doenças Transmissíveis ( DEVIT ) and by the Secretaria de Vigilância em Saúde ( SVS ) of the Ministério da Saúde . Input data for CL for these states were the place of infection at the municipality scale . In Colombia , 73 , 479 cases were spread across 882 localities in all the 32 departments of Colombia from 2007 to 2015 . Colombian data were extracted from the SIVIGILA ( National Public Health Surveillance System ) website , which gathers cases of the various diseases that require mandatory reporting . CL data were validated by the Grupo de Investigaciones Microbiológicas-UR ( GIMUR ) from Universidad del Rosario , as reported elsewhere [43] . In French Guiana the 448 cases distributed in 89 localities come from patients in consultation for suspected leishmaniasis at the LHUPM ( Laboratoire Hospitalo-Universitaire de Parasitologie et Mycologie ) and in the country’s different health centres , between 2008 and 2015 . We chose not to include cases from Venezuela , Suriname , Guyana , Bolivia and Peru , because we had no access to official cases coming from health centres that could be considered as non-biased public data . We report a geospatial analysis of CL data . For Colombia and Brazil , the data were readily obtained from existing public access databases ( Colombia: SIVIGILA , and Brazil: SINAN ) . For French Guiana , we report the cases from the database already published in Simon et al . ( 2017 ) [45] . For all data , the information that identifies the patient was anonymised in the databases and there is no need for ethical considerations . All data were processed in ArcGis 10 . 4 [46] . All variables were used at the resolution of 30 arc-seconds ( ~1 km2 ) for the Amazonian and French Guiana models . Geographic variables available at another resolution and vectorized variables were resampled at 30 arc-seconds using the nearest neighbour joining method , implemented with ArcGis 10 . 4 . The bioclimatic , environmental and anthropogenic variables are given in Table 1 , with their initial resolution . In total , 26 variables were used for the Amazonian model including 19 bioclimatic variables from WorldClim2 , three anthropogenic variables with the population density , the human poverty and the human footprint ( HFP ) and four environmental variables: the biomass aboveground , elevation , forest canopy height and species richness in mammals . For French Guiana , we used the same 19 bioclimatic variables that for the Amazonian model , plus a cloud cover variable . The same environmental variables were used as for the Amazon model with in addition , the percentage of the cell covered by high forest , the distance to river courses , the distance to forest edge and the distance to a relief at least of 500 meters . However , we did not have the species richness variable in mammals for this last model . Two anthropogenic variables were used , the density of tracks and road network and HFP; we used a specific HFP developed for French Guiana , which has a higher level of detail and a more recent update than for the Amazonian HFP variable . The detailed information and sources of the variables used for both models is available in supplementary method ( S1 Method ) .
Method 2 of the distribution of the points led to the best AUC score ( 0 . 842; 95th ranked AUC value for null model = 0 . 5073 ) ( S1 Table ) . The five variables explaining the probability of occurrence of CL cases best were , human population density ( 30 . 8% of the contribution ) , HFP ( 30 . 2% ) , Bioclim 4 ( seasonal temperature; 18 . 9% ) , mammalian species richness ( 13 . 8% ) and aboveground biomass ( 6 . 3% ) . For the jackknife test the variable with the highest gain when used alone was population density , which therefore appears to contain the most useful information by itself ( S1 Fig ) . The variable that most decreases the gain when it is omitted is Bioclim 4 , which therefore appears to have the most information that is not present in the other four variables ( S1 Fig ) . The likelihood of occurrences does not vary whatever the population density ( Fig 2A ) . The likelihood of occurrence increases sharply to a HFP value of about 50 , then decreases sharply ( Fig 2B ) . This decrease can be attributed to our method of distributing case occurrences for high HFP values , excluding the more anthropised areas and large urban centres in the Amazon ( values above 51 ) where transmission of CL is unlikely to occur given the ecology of the CL transmission cycles . The likelihood of case occurrence decreases rapidly as the seasonal temperature variation ( Bioclim 4 ) increases ( Fig 2C ) . The likelihood of occurrence of cases with mammal species richness looks like a bell-shaped curve: it abruptly increases near 110 species , since low-richness areas indicate either non-forested habitats , where CL does not occur , or disturbed forest habitats; the occurrence then decreases for the highest mammal richness values , those associated with very remote , species-rich and restricted Amazonian regions where , at least , no CL human cases are reported ( Fig 2D ) . Concerning the aboveground biomass , the likelihood of case occurrence is stable , then decreases over a very small interval of the variable , between 200 and 250 tons/ha , and finally increases when the values of the variable increase . The predicted risk map is driven mainly by population density and HFP , showing disturbed forest areas and large nuclei of human populations as foci potentially at risk for leishmaniasis transmission to human populations living in these contexts . The north-northwest of South America , mainly Venezuela , and the south-eastern part of the Amazon basin , notably near the south of the delta area , appear as the most at-risk areas for leishmaniasis transmission according to the explanatory variables retained in the models ( Fig 3 ) . The model with distribution method 2 had the best AUC ( 0 . 885 , null model = 0 . 5491 ) ( S1 Table ) . The best AUC score was obtained with four explanatory variables that included two climatic variables ( Bioclim 2 and 16; mean diurnal range of temperature and the precipitation of the wettest quarter , respectively ) , one anthropogenic variable ( HFP ) and one environmental variable ( distance to forest relief ) , with overall the most significant contribution being HFP with 70 . 1% of the total explanation . The jackknife test training shows that the explanatory variable with greatest gain when used alone and that decreases the gain the most when omitted is HFP . Jackknife analysis was performed to test the importance of each of the variables retained . Bioclim variables 2 and 16 contributed 9 . 2% and 15 . 4% of the total explanation , respectively . The last variable distance to a relief of at least 500 m seemed to contribute very little to the model ( 5 . 3% ) , but the jackknife test showed a decrease in AUC when the variable was not present in the training and the test ( S2 Fig ) . The likelihood of occurrence increases with HFP until 35–40 and then it decreases according to a bell-shaped curve . This decrease is directly related to the point distribution of method 2 since areas with HFP > 40 were excluded from contamination areas ( Fig 4A ) . For Bioclim 16 , the likelihood of occurrence slightly increases with precipitation of the wettest quarter , indicating that the occurrence of cases increases monotonically during the rainy season in this region ( Fig 4B ) and then drops for the highest values of precipitation of the wettest quarter . The response of the mean diurnal range variable ( Bioclim 2 ) shows that the likelihood of occurrence slightly decreases as the temperature amplitude increases and then sharply rises to reach a plateau for the highest values of Bioclim 2 ( Fig 4C ) . When the amplitude is the highest , there is a sharp increase in the likelihood of cases occurring , as explained by several cases of CL in the eastern part of the French Guiana region . The response curve of the distance to relief of at least 500 m variable shows that occurrence is high at 500 m and then drops off rapidly and increases gradually at lower altitudes ( Fig 4D ) . The risk map shows that prediction for CL transmission is higher where the HFP index is high , i . e . anthropogenic activities ( hunting , logging , development of activities and housing at edges ) are most common ( Fig 5 ) .
At the beginning of this study , the set of initial variables tested was large enough to encompass all the ecological complexity of the CL life cycle . In agreement with previous studies using human CL occurrence data [14 , 38 , 39] , the variable contributing most to the Amazonian model were two anthropogenic variables , i . e . population density and HFP , followed by seasonal temperature , mammal species richness and aboveground biomass . At French Guiana scale , the variables explaining the greatest number of cases were HFP , followed by precipitation in the wettest quarter ( Bioclim 2 ) and the mean diurnal range of temperature ( Bioclim 16 ) . At the Amazonian large scale , the presence of four biotic variables with wild mammal species richness , population density , HFP and aboveground biomass show the likelihood of increased case occurrence when all these parameters also increase . Several studies have shown that changes in human activities with landscape management in rural areas may affect the population dynamics and distribution of phlebotomine species in Amazonia [41 , 55 , 56] . The response of the seasonal temperature indicates that CL cases are more likely to occur in geographical areas with the least amplitude in seasonal variation . This is not surprising and lends support to the absence of CL cases in the Andes Mountains , with their present unfavourable meteorological and ecological conditions for phlebotomine vectors [33] . Although this observation is ecologically consistent for a large-scale study , it does not add information on climatic factors favouring the risk in the Amazonia biome . Here , unlike the results of previous studies [14 , 38] , the contribution of rainfall remained below 5% , probably because the model is run in the same biome where precipitation has no significant impact on the risk of CL transmission . Wild mammal species richness and aboveground biomass are reminders that the involvement of mammalian hosts and the ecology of the vector are also important biotics drivers to be considered in assessing the risk of CL [57] . Interestingly , in French Guiana , the likelihood of case occurrence is also mainly driven by the biotic HFP variable , with cases increasing as HFP rises . Although environmental policies in this region are very protective [58] , pressures on forest ecosystems have changed over the last few decades . Today , 86 . 2% of CL cases reported are due to L . guyanensis whose the life cycle is mainly sylvatic , but an increase in cases due to L . braziliensis has been observed in recent years [45] . The ecology of L . braziliensis has been assimilated with disturbed and peri-domestic forest habitats in several parts of Amazonia [37] . For this model , the HFP biotic variable probably provided a better account for anthropogenic modification on the environment given its finer resolution and more up-to-date data than those used for the entire Amazonian region [59] . For French Guiana , we observed a probability of an increase in CL case occurrence when the precipitation of the wettest quarter and mean diurnal range increased , confirming the importance of these climate variables in the Amazon basin regardless of the scale chosen . Indeed , in French Guiana a large majority of cases are in the north-east region where precipitation and mean diurnal temperature variations are the greatest . This increase can potentially be explained by the climatic conditions , which are more favourable for vector proliferation , and by the more extensive anthropogenic activities related to the forest [59] . For the response curve of the variable representing the distance to a relief of more than 500 m , the probability of cases occurring is higher on the 500-m reliefs and when one moves away from these reliefs . This result may reflect the high biological diversity of phlebotomine species with different altitudinal distributions as we observed in many regions of Southern America . Ready et al . [60] showed the presence of Psychodopygus wellcomei , the main vector of L . ( V . ) braziliensis in Amazonia , at altitudes over 500 m and then the sharp drop in the probability of occurrence of CL cases and its consistent increase can reflect the ecological requirement of vectors in French Guiana . The risk map obtained for the Amazonian model is relatively similar to the at-risk areas highlighted by a previous study at the South American scale [14] . However , it differs from the map obtained by Purse et al . [38] where the entire Amazon basin was found at risk . In the present study , the AUC , omission test and the null model suggest that the predictions are reliable . The predominantly identified risk areas are described where the human impact on the environment is substantial , i . e . , close to urban centres and along roads and rivers where human populations are concentrated . Venezuela , north-east of Brazil , and northern Bolivia emerge as potential at-risk areas while no case of CL in this region was used in the model . The data currently available on CL indicate that cases have been identified in these areas [61] , although they are not being included as input data , suggesting that the model did not make a significant commission error . In Colombia , the states in the south-east did not come out as a potentially high-risk area . This result seems to contradict the recent study conducted by Herrera et al . [43] , which indicated that these states had the highest incidence and number of cases in the country . This failure may be explained by the limit of spatial ENM when working with quantitative data . Despite a very high number of reported CL cases in this region , the number of cities and the population density remain very low . However , ENMs handle quantitative data such as prevalence , because the information is retained at the pixel scale and whatever the number of cases in one pixel , it is saturated with the first reported case . Despite our procedure to create a buffer zone to randomly disperse cases , cases and substantially increase the number of available pixels to distribute the cases , the model still gives greater importance to areas where the spatial occurrence of cases is widely distributed . For French Guiana , this is the first study to propose a high-resolution risk map based on precisely geolocalised cases . For this European territory , high-risk areas are located where the anthropogenic pressures on habitats are the strongest . A risk zone appears on the map in the west of this region despite the absence of cases , suggesting under-reported and/or under-diagnosed cases . French Guiana is a region where deforestation , hunting , forestry activities , and legal and illegal gold panning have increased in recent years [62] . This information , collected on the importance of the influence of human activities in increasing the risk of this disease , as well as the numerous studies carried out on the possible anthropisation of the vector cycle as shown in Colombia [34] and Manaus , Brazil [63] , suggest that human activities in the rainforest in the Amazon and French Guiana could promote a peri-domestication of the CL disease cycle . Also , throughout Amazonia , people could be infected in peri-urban forest fragments with great canopy cover , which is essential for maintenance of the Leishmania vector/reservoir species diversity and abundance [64–66] . The methodology of this study is based on satellite imagery and correlative analyses , but it remains a visual assessment . It also excludes that the cycles could occur in anthropised and highly disturbed habitats . Indeed , in Colombia CL is linked to the urban cycle [34] and in the largest Amazonian cities such as Belém , CL is associated with small forest fragments surrounded by an urban area and where ( phlebotomine ) putative vectors may sustain [64] . Consequently , it may be interesting to retain relatively high values of HFP in order not to completely obscure the likelihood of local peri-domestication of CL . Another limitation of our study is that some areas of the Amazon biome are not considered at risk while we do know the existence of CL cases , as in Peru and Bolivia . Heterogeneity in the availability of our data increases the models’ omission rate , but we favoured data that were reliable and retrieved directly from the public health database for each country . Unfortunately , it was possible to find this kind of data for only two countries , i . e . , Colombia and Brazil , and for the French Guiana region . We also attempted to obtain the most updated variables for the Amazonian model , but some are not updated over the period when the cases occurred , so the environmental data are not necessarily concomitant with the case occurrence period . In addition , we are aware that the models are highly dependent on the input variables and spatial scaling , so risk maps produced with large-scale data and models should not be extrapolated for more restricted geographical areas; risk maps are first context- and space-dependent . Modelling a parasite system that is based on several species of hosts and reservoirs requires considering relevant biotic and abiotic variables summarising the ecological conditions in which the transmission cycle takes place . For this complex issue , the BAM diagram may help to select the variables and the scale of study . Finally , for both models ( Amazonia and French Guiana ) this study highlighted the importance of considering the anthropogenic drivers for risk assessment . This conclusion differs from that proposed by Pigott and collaborators , [14] who argued that climatic conditions were the main driver of CL case distribution in South America . The adequate choice of the spatial scale under scrutiny , in accordance with the variables explored , can be a major determinant in the discrepancy that we observed between Pigott et al . and our present results . Therefore , risk mapping should not be made without considering variables representing the vulnerability of human individuals and communities to the disease and further add to the importance of an appropriate scaling when designing ENM studies [50 , 67] . Generally , coarse-scale studies appear to favour the importance of climatic variables in explaining infectious disease presence and spatial distribution [68] . This pattern has already been referred to as Eltonian Noise Hypothesis [69] which assumes that local biological interactions or microhabitat biotic conditions required by a specific parasite cycle should not affect niche estimates at coarse scales [19] . Many studies have attempted to make future projections of climate change on vector-borne diseases to determine the factors favouring disease emergence and to predict the dispersal of infectious disease agents . For diseases whose transmission cycles are confined to restricted geographic areas , it is likely that the small-scale human impact firstly may influence spatial expansion or regression of these diseases . With the methodological framework proposed here and with fine-scale and updated variables on anthropogenic disturbances , ENMs remain a valuable tool to determine local factors that are the drivers of parasite transmission and may help relevant decision-making by health authorities . Every ENM study that uses risk modelling should target the proper scale based on these elements . This statement can be extended most particularly to the Leishmania ecological system . In French Guiana , the CL system is mainly represented by L . guyanensis and Nyssomyia umbratilis with Xenarthran species acting as major host reservoirs [31 , 45] , while this cannot be identical for other pan-Amazon regions with other species involved in the cycle [34 , 70] . The relevance of developing future models of CL risks with only climatic variables is questionable . Indeed , it is likely that the policy and economic decisions with their cascading impacts on poverty , hygiene , war , displacement of populations , etc . , and short-term local planning strategies [71] will have a more direct and immediate impact on biodiversity and their interactions with disease components . This is particularly true in regions where the expected climatic variations will remain low compared to the impact of microclimates created , for example , by the creation of hydroelectric dams [40] , the burden of extensive agriculture [72] or the effects of edge habitats [73] . These anthropogenic factors will remain extremely difficult to control in the future and will continue to challenge the relevance of predictive models , whatever the ongoing methodological improvements and the quality of the data used as independent variables in models . | Cutaneous leishmaniasis is a vector-borne zoonotic disease with a complex transmission cycle that includes many parasite , vector and host species . This disease continues to pose public health problems worldwide despite the measures put in place . In recent years , methodological tools commonly used in ecology , called ecological niche prediction models , have made it possible to determine the environmental and anthropogenic variables that may be favourable to the presence of the host and vector species communities involved in the cycle and therefore to the presence of certain disease agents . The use of these models , based on the presence of human cases of the disease , can overcome some of the uncertainties concerning the diversity of the vectors and the potential hosts involved in the transmission cycle . This approach of health ecology combining ecology and epidemiology could provide new insights into understanding the cycle of disease transmission and the influence of environmental factors and thus improve the prediction of disease emergence and epidemics . It can be applied to various vector-borne diseases whose transmission cycles are still poorly understood and for which studies classically carried out in epidemiology have not prevented disease progression . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"forest",
"ecology",
"tropical",
"diseases",
"vertebrates",
"parasitic",
"diseases",
"animals",
"mammals",
"neglected",
"tropical",
"diseases",
"population",
"biology",
"ecological",
... | 2019 | Ecological niche modelling for predicting the risk of cutaneous leishmaniasis in the Neotropical moist forest biome |
Cells in the wing blade of Drosophila melanogaster exhibit an in-plane polarization causing distal orientation of hairs . Establishment of the Planar Cell Polarity ( PCP ) involves intercellular interactions as well as a global orienting signal . Many of the genetic and molecular components underlying this process have been experimentally identified and a recently advanced system-level model has suggested that the observed mutant phenotypes can be understood in terms of intercellular interactions involving asymmetric localization of membrane bound proteins . Among key open questions in understanding the emergence of ordered polarization is the effect of stochasticity and the role of the global orienting signal . These issues relate closely to our understanding of ferromagnetism in physical systems . Here we pursue this analogy to understand the emergence of PCP order . To this end we develop a semi-phenomenological representation of the underlying molecular processes and define a “phase diagram” of the model which provides a global view of the dependence of the phenotype on parameters . We show that the dynamics of PCP has two regimes: rapid growth in the amplitude of local polarization followed by a slower process of alignment which progresses from small to large scales . We discuss the response of the tissue to various types of orienting signals and show that global PCP order can be achieved with a weak orienting signal provided that it acts during the early phase of the process . Finally we define and discuss some of the experimental predictions of the model .
Epithelia in diverse tissues , in addition to their apico-basal polarization , acquire a polarization within the two-dimensional layer of cells – a phenomenon called planar cell polarity ( PCP ) [1]–[5] . In the developing wing of Drosophila , PCP determines the growth direction of small hairs that extend radially from cell boundaries . In a wild-type wing , where cells are approximately hexagonal and form a regular honeycomb lattice , all of these hairs point to the distal direction . A series of recent experiments show that several key proteins [6] , including the transmembrane proteins Frizzled ( Fz ) and Van-Gogh ( Vang ) and the cytosolic proteins Dishevelled ( Dsh ) and Prickled ( Pk ) , localize asymmetrically on cell boundaries [7]–[12] - defining a direction in the plane within each cell and forming a characteristic zig-zag pattern of protein localization on the lattice ( Fig . 1A ) . Other experiments show that local PCP orientation depends on inter-cellular signaling . First , mutant clones in which fz or Vang activity is suppressed or amplified , cause characteristic and reproducible inversion of polarity in large patches of cells that are proximal or distal to the clone [13] . These observations are summarized in Figs . 1 C , D . Second , in fat mutant clones [14] , [15] hairs do not all point correctly in the distal direction , yet , their orientation is strongly correlated between nearby cells and varies gradually across the tissue creating a characteristic swirling pattern . Thus the experimental evidence suggests that an interaction between neighboring cells tends to locally align their polarity [1] , [3] , [14] . This local polarity need not point distally unless , in addition , there is a global orienting signal that picks out the distal direction throughout the wing ( most likely originating with the Dpp morphogen gradient which defines the Anterior-Posterior axis of the wing in the larval stage of development [16] ) . Yet , aside from a clear involvement of protocadherin fat [17] , [18] the molecular details of this pathway remains for now unknown . The swirling patterns in fat mutants [14] and recent evidence [15] , [19] , suggest that the orienting field is related to the presence of a “gradient” in the fat , four-jointed , and dachs pathway . These observations evoke an analogy between PCP and the behavior of ferromagnets , extensively studied in physics and well understood in terms of statistical mechanics of relatively simple models [20] . In these models each atomic site is assigned a magnetic dipole – spin – which can assume a different orientation ( analogous to the direction of polarization in an epithelial cell ) . The salient properties of ferromagnets arise from the opposing influence of an interaction between neighboring spins , which tends to co-align their orientation , and the influence of thermal fluctuations , which tend to randomize the spin direction . Ferromagnets typically exhibit two phases of behavior: a high temperature phase , where spins are disordered and a low temperature ferromagnetic phase , where the interactions dominate over thermal fluctuations – leading to a spontaneous polarization in an arbitrary direction . In this state even a small external magnetic field has a big effect on magnetic polarization as the spontaneous polarization aligns itself with the external field , yet the dynamics leading to global alignment can be quite slow . An essential lesson from statistical mechanics is that the ordered and disordered states exist in a broad class of models and can be discussed in a general context , focusing on a classification of the different regimes as a function of a few parameters . We follow this lesson by focusing the study on the competition between the intercellular interaction and the disordering influence of the fluctuations introduced by the noisy molecular interactions . As in statistical mechanics we define a phase diagram which identifies different regimes of behavior in the space of the most relevant parameters . We then address the role of the global directional signal in the dynamics of global alignment . A molecular model for PCP formation was recently proposed in Ref . [21] , and was shown to reproduce a number of experimental findings . This model involves 38 parameters that were adjusted to successfully reproduce a set of wild-type and mutant phenotypes . Here we pursue an alternative approach and instead of moving on to more and more complex models develop a model with a smaller number of degrees of freedom and a smaller number of parameters . Instead of fixing a particular set of parameters by fitting the data we explore the generic behavior of the model as a function of parameters defining quantitative features characteristic of the different phases . In formulating the model we identify several essential ingredients , required to obtain the characteristic zig-zag pattern and the non-autonomy of fz and Vang mutant clones . We expect our simplified model to capture important properties of PCP , although it does not incorporate all the molecular details . After discussing the essential ingredients of the model , we obtain a phase diagram describing its steady state properties . We then consider the dynamics of local polarization strength and orientation in the absence and in the presence of a global orienting signal . We show that global alignment can be achieved with a weak global orienting signal provided it is present throughout the tissue at the earliest stage of PCP dynamics . Finally we discuss the experimental predictions coming out of the model and the tools required to test these predictions .
Three essential ingredients are included in the model , to account for the characteristic zig-zag patterns of protein localization and for the non-autonomy of fz and Vang mutant clones . There are several reasons why the dynamic equations are not deterministic . Even in the steady state , interfacial complexes not only bind and unbind due to thermal fluctuations , but like nearly everything else inside the cell are being constantly recycled and reassembled . Stochastic fluctuations arise from the molecular noise of reactions and the variability in the state of the cell defining the “intrinsic” and “extrinsic” noise [22] . It will suffice however to describe stochasticity of complex binding and unbinding as if it were a Poisson process . Equation ( 1 ) is thus replaced by a stochastic equation , ( 4 ) [and a similar modification applies to Eq . ( 2 ) ] where the noise can be approximated as white Gaussian noise if the number of molecules per cell is not too small . Assuming that the dominant contribution comes from the finite number of molecules participating in the binding/unbinding dynamics , the variance of is inversely proportional to ( see Methods ) , where is defined as the number of molecules per interface: where is the total concentration of molecules ( bound and unbound ) and is the area of an interface ( about – see Fig . 1B ) . Since the variance of decreases with increase of , plays a role similar to temperature in a ferromagnet . If there are Fz molecules per cell [23] , is of order resulting in the root-mean-square fluctuations of the order of ( i . e . ) of the mean . Other sources of intrinsic noise , in addition to the stochasticity of binding and unbinding events , may increase the noise variance beyond the above estimate . These additional noise sources include , for example , stochasticity in the signaling pathway that generates the non-local inhibition within each cell , or fluctuations in and . Such sources of intrinsic noise , acting upstream of and , are propagated to the PCP signaling dynamics through the dynamics of complex formation , and can thus be described qualitatively by the noise term in Eq . ( 4 ) , with an effective value of that is possibly smaller than predicted from the number of and molecules alone . What are the consequences of the model defined above when cells are arranged on a hexagonal lattice ? Let us first consider the steady state in the deterministic limit . Fig . 3A shows a typical phase diagram on a two-dimensional plane dissecting our five dimensional parameter space ( see Methods ) : the axis is the range of the non-local interaction in units of the cell lattice spacing , and the axis the coefficient which controls inhibition ( see Methods ) . In the region labeled there is a unique steady state in which there is no polarization of the protein distribution . In contrast , in region the stable steady state has the symmetry shown in Fig . 3B: Both and distributions carry a vector dipole moment that points towards the center of a side , and due to the lattice symmetry there are six equivalent states of this type . A uniform steady state exists as well , but it is unstable . Region differs from in the direction of the dipole , which points towards a vertex instead of pointing towards and edge ( Fig . 3C ) . The transition from the uniform state , , to the edge state , in the phase diagram is continuous: the dipole moment tends to zero when approaching the phase boundary from the side . A similar transition from a state to a state can exist as well , and is present on another two dimensional “slice” through the parameter space of our model . This transition is also continuous . We next consider the effects of stochasticity , which were ignored in the discussion above by setting . When is finite ( similar to a non-vanishing temperature in a spin model ) , we ask whether the steady state maintains long-range order: i . e . whether a particular orientation is singled out throughout the lattice and the dipole moment has a non-zero average . In the language of the analogy with magnetic systems this would be a ferromagnetic state . The latter disappears as the temperature increases above a certain critical value , giving way to a paramagnetic state where dipole moments point in random directions and the average polarization vanishes ( an intermediate state with quasi-long range order may exist as well , in similarity to 2-dimensional clock models [24]–[27] ) . Hence , we expect an ordered state to be stable only when is sufficiently large , and this is indeed observed in our simulations ( Fig . 4 ) . Yet with a realistic number of molecules per cell , in the order of several thousands , the vertex and side states in our model are typically ferromagnetic . It may thus appear that when takes realistic values the system is in an ordered state and stochasticity is altogether unimportant . However , as we discuss next , the steady state is not necessarily reached within the time scales of wing development , and stochasticity plays an important role in the dynamics of ordering . Let us consider the dynamics of PCP formation , first in the absence of a global orienting signal . Fig . 5 shows results from a stochastic simulation , starting from a state where and are uniformly distributed in all cells . We can identify two stages of the process . The first stage corresponds to a gradual build up of a dipolar polarization on the cellular level . The dipole initially points in a random direction , but as its amplitude increases with time ( Fig . 5B ) local polarization begins to re-orient . At the end of this stage , when amplitude saturates , there is no global choice of PCP direction , but the orientation of nearby cells is strongly correlated: as an example , Fig . 5A shows the configuration of dipoles shortly after saturation . The second stage , which follows amplitude saturation , exhibits slow coarsening dynamics [27]: polarity direction is approximately aligned within discrete domains , the size of which gradually expands by movement of their boundaries . Note also the existence of vortex-like defects [28] ( Fig . 5A and Fig . S1 ) . Coarsening ultimately leads to a spatially uniform steady state , but this process occurs over a long time scale compared to that of amplitude growth . A quantitative theory of the early dynamics is obtained from the linear instability of the uniform steady state ( described in detail in Text S1 , part II ) . The variance of the local dipole amplitude increases exponentially in time with a characteristic time scale , ( 5 ) where for simplicity numeric prefactors of order unity are omitted ( see Text S1 , part II ) . In this equation is the amplitude of noise in the unstable uniform steady state , and both and are found from the instability analysis ( Text S1 , part II ) . This prediction is shown in Fig . 5B ( dashed line ) for comparison with the simulation . Two additional insights come from the analysis of early dynamics ( Text S1 , part II ) . First , PCP is initially isotropic , despite the discrete 6-fold symmetry of the hexagonal cell lattice . Consequently , the dipole moment initially has equal probability to point in any direction in the interval . Second , the spatial correlation established during the early dynamics typically has a longer range in the direction parallel to the dipole , compared to the perpendicular direction . These two properties of the dynamics lead to a characteristic swirling pattern before non-linearities set in . The range of correlation at this stage depends on the location in the phase diagram and increases logarithmically as a function of . We next consider how various types of symmetry-breaking orienting signals influence PCP dynamics . | Epithelial tissues are often polarized in a preferred direction which determines , for example , the direction of hair growth on mammalian skin , the orientation of scales in fish , the alignment of ommatidia in the fly eye and of sensory hair cells in the vertebrate cochlea . This in-plane polarization , known as planar cell polarity , is one of the morphogenetic fields that play a role in tissue patterning during development . Here we focus on planar cell polarity in the fly wing , where protein localization and inter-cellular ligand-receptor interactions combine with an unknown orienting signal to establish planar cell polarity of the wing epithelium . We demonstrate an analogy between this process and models of ferromagnetism in physical systems that have been studied extensively using the tools of statistical mechanics . The analogy helps in understanding how local interactions between cells can lead to global polarization order and elucidate the role of global orienting signals and the dependence of the dynamics of the process on parameters . We demonstrate that in the absence of an external orienting signal swirling patterns should emerge due to random noise . We propose ways to test this prediction and ways to quantify the magnitude and spatial variation of the unknown external orienting signal . | [
"Abstract",
"Introduction",
"Results"
] | [
"physics",
"developmental",
"biology/morphogenesis",
"and",
"cell",
"biology",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"biophysics/theory",
"and",
"simulation",
"cell",
"biology/developmental",
"molecular",
"mechanisms",
"computational",
"biology/systems",
... | 2009 | Order and Stochastic Dynamics in Drosophila Planar Cell Polarity |
Homeotic genes code for key transcription factors ( HOX-TFs ) that pattern the animal body plan . During embryonic development , Hox genes are expressed in overlapping patterns and function in a partially redundant manner . In vitro biochemical screens probing the HOX-TF sequence specificity revealed largely overlapping sequence preferences , indicating that co-factors might modulate the biological function of HOX-TFs . However , due to their overlapping expression pattern , high protein homology , and insufficiently specific antibodies , little is known about their genome-wide binding preferences . In order to overcome this problem , we virally expressed tagged versions of limb-expressed posterior HOX genes ( HOXA9-13 , and HOXD9-13 ) in primary chicken mesenchymal limb progenitor cells ( micromass ) . We determined the effect of each HOX-TF on cellular differentiation ( chondrogenesis ) and gene expression and found that groups of HOX-TFs induce distinct regulatory programs . We used ChIP-seq to determine their individual genome-wide binding profiles and identified between 12 , 721 and 28 , 572 binding sites for each of the nine HOX-TFs . Principal Component Analysis ( PCA ) of binding profiles revealed that the HOX-TFs are clustered in two subgroups ( Group 1: HOXA/D9 , HOXA/D10 , HOXD12 , and HOXA13 and Group 2: HOXA/D11 and HOXD13 ) , which are characterized by differences in their sequence specificity and by the presence of cofactor motifs . Specifically , we identified CTCF binding sites in Group 1 , indicating that this subgroup of HOX-proteins cooperates with CTCF . We confirmed this interaction by an independent biological assay ( Proximity Ligation Assay ) and demonstrated that CTCF is a novel HOX cofactor that specifically associates with Group 1 HOX-TFs , pointing towards a possible interplay between HOX-TFs and chromatin architecture .
The homeotic genes ( Hox genes ) are key regulators of development . They encode homeodomain transcription factors ( HOX-TFs ) that are expressed in an overlapping fashion along the anterior-posterior axis in all metazoans [1] . In the vertebrate genome , Hox genes are organized in clusters with their order reflecting not only their expression along the anterio-posterior body axis but also their temporal expression onset ( spatio-temporal collinearity ) . In most vertebrates , two rounds of whole-genome duplication have resulted in four clusters of Hox genes , coding for a total of 39 HOX-TFs . All HOX-TFs show high levels of sequence conservation between paralog groups ( e . g . HOXA9 and HOXD9 ) and to a lesser extent between genes of the same cluster ( e . g . HOXA1 to HOXA13 ) [reviewed in 2 , 3] . In the developing vertebrate limb , the posterior genes of the HoxA and HoxD clusters ( Hox9-13 ) are expressed along the proximo-distal axis following a collinear strategy [4] . Genetic experiments inactivating individual Hox genes revealed a remarkable redundancy within paralog groups controlling the development of the proximal ( stylopod ) , middle ( zeugopod ) , and distal ( autopod ) parts of the limb [5 , 6] . For example , neither the homozygous deletion of Hoxa11 nor Hoxd11 in mice leads to substantial malformations of the stylo- , or zeugopods . However , deletion of both Hoxa11 and Hoxd11 causes a severe truncation of the stylopod and loss of the zeugopod [7 , 8] . A similar redundancy is observed between genes of the same cluster . Deletions , in mice , that encompass the entire Hoxd13 gene cause the adjacent Hoxd12 to be expressed in a Hoxd13-like pattern associated with the functional rescue of the Hoxd13 deficiency . A similar deletion , removing Hoxd13 and Hoxd12 causes Hoxd11 to be expressed in a Hoxd13-like pattern; however , Hoxd11 is not able to rescue the loss of its two adjacent paralogs [9] . In spite of the insights gained by these elegant series of genetic experiments , the high degree of HOX protein similarity and the overlap of expression domains have hindered the elucidation of the individual HOX-TF functions . HOX-TFs were also included in large biochemical surveys to identify the specific binding sequence of transcription factors [10–12] . Two complementary studies applying protein binding microarrays ( PBM ) and SELEX-seq on purified DNA-binding domains demonstrated that all posterior HOX-TFs bind to similar AT-rich sequences that vary in their 5’ region but share a characteristic TAAA sequence in their 3’ half . Moreover , two NMR-based studies showed binding of HOXA13 to the HOXD13 site and vice versa [13 , 14] . Thus , the DNA binding specificity is not sufficient to explain individual HOX-TF function . More recent studies revealed a crucial role for cofactors in HOX-TF specificity . HOX-cofactors were shown to specifically alter the recognition sequence of the HOX-TFs by forming heterodimers [10 , 15 , 16] . Moreover , the analysis of HOX-cofactor specific binding sites suggested that these altered binding sites might be functionally more relevant for HOX binding than the HOX-TFs binding sites themselves [17] . However , due to high sequence homology , inadequate antibody specificity , and overlapping expression patterns little is known about genomic binding of the different HOX-TFs and how this might relate to their biological function . Here , we have analyzed and systematically compared the effects of nine limb bud-expressed HOX-TFs ( HOXA9-13 and HOXD9-13 ) on cell differentiation and gene regulation and compare their genome-wide binding characteristics . To mimic the natural HOX environment as closely as possible , we used mesenchymal chicken limb bud cells and moderate retroviral overexpression [18] . In this primary cell culture system ( chicken micromass , chMM ) the cells normally undergo chondrogenic differentiation; a process that can be altered by virally expressed transgenes [18] . Given the identical cell origin , culture conditions , and antibody use , this system allowed us to assess the distinctive properties of each HOX-TF and compare them to each other . We find that certain HOXA/HOXD paralog TFs have opposing effects on chondrogenic differentiation and induce distinct regulatory programs in transduced cells . Further , by comparing the genome-wide DNA binding of nine HOX-TFs in this experimental setting , we find that the posterior HOX-TFs can be separated into two groups ( Group 1 and Group 2 ) , with distinct binding motifs and distinct associations with cofactors . Finally , we characterized CTCF ( the CCCTC-binding factor ) as a novel cofactor of HOX-TFs and show that Group 1 but not Group 2 HOX-TFs bind thousands of CTCF-occupied sites in the chicken genome .
To systematically compare the function of posterior HOX-TFs , we virally expressed FLAG-tagged versions of each TF in chicken micromass ( chMM ) cultures . After validating a reproducible , similar , and moderate expression of virally expressed HOX-TFs ( S1 Fig ) , we assessed the effect induced by the different HOX-TFs on chMM cultures . We noticed that some HOX-TFs promoted chondrogenic differentiation ( HOXA9 , HOXA10 , HOXD10 ) , while others inhibited the process ( HOXD9 , HOXD11 , HOXA11 , HOXD12 , HOXA13 , and HOXD13 ) ( Fig 1A ) . Interestingly , paralogue HOX-TFs did not always have the same general impact on the chondrogenic differentiation of the chMM . While HOXA9 stimulated chondrogenic differentiation , its paralog HOXD9 inhibited the same process . In contrast , HOXA10 and HOXD10 both promoted chondrogenic differentiation . HOXA11 and HOXD11 both inhibited chondrogenic differentiation , but to a very different extent . Finally , HOXD13 and HOXA13 both strongly inhibited cartilage formation; however , Eosin staining showed that the cell morphology of the HOXA13-expressing chMM was quite distinct from HOXD12 or HOXD13 cultures ( Fig 1A ) . The simple readout of the chMM morphology showed that the HOX-TFs induce distinct effects on cell differentiation . In order to comprehensively compare the effects on gene expression , we performed RNA-seq of HOX-TF expressing chMM cultures . We used DEseq2 [19] to generate a list of genes that were differentially regulated compared with mock-infected chMM cultures . We then used the genes that were found among the 50 most strongly regulated genes in any of the nine datasets for hierarchical clustering ( Fig 1B , S1 Table ) . The hierarchical clustering recapitulated some of the main differences found between HOX-TFs that were detected in chMM gross morphology . HOX10 and HOX11 paralogs clustered together , while HOX9 paralogs , which bore striking differences in chMM morphology , clustered apart . Furthermore , the clustering process classified the paralog groups in an order that partially corresponded to their known role in limb development . The clustering separated the stylo-/zeugopod expressed HOX-TFs ( HOXD9 , HOXA/D10 , HOXA/D11 ) from the autopod expressed HOXD12/13 . Two factors , HOXA9 and HOXA13 , clustered separately from all other HOX-TFs . This indicates that despite comparable effects on chMM morphology , the regulatory programs induced by HOXA9 and HOXA13 are distinct from the other posterior HOX-TFs . Moreover , the HOX11 paralogs induced transcriptional programs so similar to one another that the clustering algorithm was not able to separate the two replicate datasets from each factor . Interestingly , two genes coding for subunits of the AP1 class of transcription factors , JUN and FOS , were among the most strongly upregulated genes in all of the datasets , suggesting that they might be direct targets of HOX-TFs . Finally , we assessed whether the expression of any HOX-TF had a regulatory effect on the HOXA/D gene cluster , since the genetic experiments had suggested the possibility of HOX cross-regulation . In fact , we find that all posterior HOX-TFs , with the exception of HOXA9 , where able to induce HOXA/D13 expression at least two fold in comparison to mock-infected chMM cultures . In contrast , HOXA9 had generally a repressive effect on the other HOX genes , especially toward its genomic neighbours in cis , HOXA10 and HOXA11 ( S2 Fig ) . Taken together , our analysis shows that , despite high homology and functional redundancy in vivo , the direct effects of paralog HOX-TFs in chMM cultures are distinct . While some can be similar ( HOXA10/D10 and HOXA11/D11 ) others can have opposing effects ( HOXA13/HOXD13 and HOXA9/D9 ) . We next wanted to assess whether analogous differences could be observed between paralog groups in their genome-wide binding preferences . We generated ChIP-seq profiles from two biological replicates of virally expressed HOX-TFs in chMM cultures using the αFLAG antibody . We identified between 12 , 721 and 28 , 572 binding sites for each of the nine HOX-TFs ( Fig 2 , S1E and S1F Fig ) . We first assessed the binding sites shared between HOX-TFs from the same paralog groups by taking the 10 , 000 strongest peaks for each factor and calculated the pairwise overlap between all HOX-TFs . Similar to the results of the expression analysis , the HOX10 and HOX11 paralogs shared more peaks ( 78–81% and 85–86% , respectively ) than the HOX9 and HOX13 paralogs ( 65–60% and 29–19% ) ( S4A Fig ) . We then mapped location of HOX-TF binding sites relative to genes . The genomic binding of HOX-TFs respective to genes was similar for all nine factors . The majority of binding sites were either intronic ( 17–33% ) or intergenic ( 53–60% ) and only 9–21% of the peaks were located in annotated promoters ( -5kb to +2kb of known TSS ) . Next , we performed a principal components analysis ( PCA ) to compare the datasets in an unbiased way , using the identified peaks as input ( Fig 2B ) . PCA showed that the binding of HOX-TF paralogs seemed to be more similar than their effects on chMM differentiation and gene expression . HOXA13 and HOXD13 were a notable exception as they clustered separately from the other HOX-TFs along PC1 ( Fig 2B , dashed box ) . In addition , they were also very different from one another in PC2 . A comparison of all tested HOX-TFs in PC2 revealed a surprising separation into two groups , which neither reflected the effects on cell differentiation and gene expression , nor the sequence homology of the TFs . Group 1 comprised HOXA/D9 , HOXA/D10 , HOXD12 , and HOXA13 ( Fig 2B , blue ) and Group 2 comprised HOXA/D11 and HOXD13 ( Fig 2B , black ) . To find a possible cause for this separation , we first tested whether the grouping could be attributed to the sequence-specificity of the TFs . For this we performed de novo motif analysis using the peak-motifs algorithm [20] with the 5 , 000 strongest peaks as input and compared it to the published results from PBM and SELEX-seq ( Fig 2C and S4B Fig ) . This comparison showed a general similarity between in vitro and ChIP-seq derived motifs . However , several sequence features had not been detected in the previously published datasets . We found a prominent G at the 5’ end of all Group 1 motifs ( HOXA/D9 , HOXA/D10 , HOXD12 , and HOXA13 ) , which had also been detected using SELEX-seq [12] . More striking , we found that the TAAA 3’ end , which is a characteristic of posterior HOX-TFs , changed to a TGAAA in all Group 1 HOX-TF motifs , with the notable exception of HOXA13 . The motifs identified for HOXA13 and HOXD13 were identical to the ones detected in PBM/SELEX-seq . In contrast , the primary motif of HOXA11 and HOXD11 did not overlap with those detected in the corresponding in vitro datasets . Specifically , the CCATAAA motif ( HOXA/D11 ) we observed was highly similar to a change in sequence specificity that HOXA10 undergoes when co-binding with PBX4 [10] . Generally , motif analysis for the HOX-TFs identified not only primary motifs but also several alternative HOX-like motifs , suggesting that the DNA-dependent binding of HOX-TFs might be less sequence-driven than other TFs ( S4B Fig ) . Group 1 and Group 2 HOX-TFs also revealed differences , when we considered the fraction of ChIP-seq peaks that contained a HOX-TFs binding site ( Fig 2D and S4C Fig ) . The number of peaks carrying a HOX-binding site ( i . e . matching one of the top three HOX motifs ) was relatively low in general , ranging from as little as 15% ( HOXD9 ) to 44% ( HOXA13 ) . Interestingly , the three Group 2 HOX-TFs had all high numbers of HOX binding sites in contrast to the Group 1 HOX-TFs , which displayed the lowest number of peaks carrying HOX-TF binding sites . As in the PCA and de novo motif analyses , HOXA13 was a notable exception of the Group 1 HOX-TFs , as it contained a high number of peaks carrying binding sites . To exclude the effect of weak and maybe indirect binding sites from the analysis , we performed the same analysis for the 10 , 000 and 1 , 000 strongest peaks ( S4C Fig ) . Although the fraction of binding site-containing peaks slightly increased , the general distribution stayed the same . The relatively low numbers of HOX-TF peaks containing HOX binding sites indicated that other factors might contribute to DNA binding . Sequence analysis of ChIP-seq peaks allows not only for the detection of sequence-specific binding sites but also for the identification of putative cofactors . Therefore , we performed a de novo motif analysis using all peaks as input and then compared the non-HOX like motifs to the literature and to large TF motif databases ( JASPAR [21] , footprint DB [22] ) ( Fig 3 ) . In Group 2 of HOX-TFs , we were not able to detect any clear cofactor motif . In contrast , we found three putative cofactor motifs in five out of the six Group 1 HOX-TF peak sets . The first motif was the well-characterized TGANTCA AP1 binding site [23] ( Fig 3A ) . A second motif , CGCTCCC/G was detected with high specificity in the HOXA9 and HOXD9 peaks and with lower specificity ( but still among the top 5 ) in the HOXA10 , HOXD10 and HOXD12 peaks ( S5A and S5B Fig ) . This motif was particularly enriched in HOXA13 peaks ( S8 Fig ) . We were not able to find matching or similar motifs in the JASPAR and footprint-DB databases , raising the possibility that it either represented the binding site of an uncharacterized TF or a composite binding site recognized by a dimerized TF complex . As a third motif , we detected a 12bp long GC-rich motif in all Group 1 HOX-TF datasets except HOXA13 . This motif perfectly matched the known motif of the CCCTC-binding factor ( CTCF ) , a well described TF involved in gene regulation and genome architecture ( Fig 3A and S5 Fig ) [24] . The de novo discovery of cofactor motifs can be masked by the strong overrepresentation of the primary motif . To exclude this possibility , we performed a reverse search and identified and counted all matches to CTCF ( Fig 3 , S6 Fig ) or AP1 ( S7 Fig ) binding sites in the nine HOX-TF data sets . For the CTCF binding sites , this reverse search revealed a characteristic difference between Group 1 and Group 2 HOX-TFs . Altogether , 14–23% of all Group 1 HOX-TF peaks , but only 5–9% of Group 2 HOX-TF peaks contained a CTCF binding site ( S6A Fig ) . In contrast , we identified AP1 binding sites in about 3–7% of all peaks of the different HOX-TFs and there seemed to be no distinction between Group 1 and Group 2 HOX-TFs ( S7 Fig ) . Next , we mapped the position of the CTCF binding sites within the HOX-TF peaks and found that in Group 1 , but not Group 2 , the CTCF sites were located predominantly near the peak summits ( Fig 3B and S6B Fig ) , suggesting a binding mode in which the HOX-TF binds indirectly via CTCF . This was further supported by a discriminatory motif analysis , which revealed that Group 1 HOX-TF peaks contained either a HOX or a CTCF binding site and that only a minority of HOX-TF peaks contained binding sites for both TFs ( Fig 3C ) . Motif analysis indicated that CTCF and Group 1 HOX-TFs might co-bind to many sites throughout the genome . We , therefore , mapped CTCF binding sites genome-wide by virally expressing FLAG-tagged CTCF in chMM cultures ( Fig 4 ) and performed ChIP-seq using the αFLAG antibody . From the same sample , we also performed ChIP-seq for endogenous RAD21 , a subunit of the cohesin complex and an important CTCF-cofactor [25] . We identified 22 , 357 CTCF and 17 , 589 RAD21 binding sites . Similar to previous reports , CTCF and RAD21 co-bound to 53% of all CTCF and to 67% of all RAD21 peaks . We then tested how many HOX-TF peaks overlapped with CTCF or RAD21 peaks . We observed that the characteristic distinction between Group 1 and Group 2 HOX-TFs could be recapitulated at ChIP-seq binding sites . Indeed , Group 1 HOX-TFs shared between 15% and 24% of their peaks with CTCF ( 12–20% with RAD21 ) , whereas only 3–8% of Group 2 peaks overlapped with CTCF ( 3–7% with RAD21 ) ( Fig 4B and S9A , S9B and S9C Fig ) . Finally , we set out to find differences between CTCF-HOX shared or HOX-only binding sites using HOXA10 ( Group 1 ) as a representative example . We first tested whether the CTCF-HOXA10 co-bound sites were enriched for promoters or intergenic regions but found no differences between HOXA10-only , HOXA10-CTCF shared or CTCF-only peaks ( S10 Fig ) . We then looked for underlying binding sequences in the 24% of HOXA10 peaks that are shared with CTCF and observed that 69% of them contained a CTCF binding site ( 23% in all HOXA10 peaks ) . In contrast , only 16% of the peaks had a HOXA10 binding site ( 18% in all HOXA10 peaks ) , suggesting that HOXA10 indirectly binds to these CTCF-shared peaks via CTCF . Taken together , motif analysis of HOX-TF binding sites and ChIP-seq for CTCF/RAD21 both found Group 1 , but not Group 2 HOX-TF binding associated with CTCF/cohesin ( Fig 4C ) . Both , motif analysis and peak overlap strongly suggested an interaction between Group 1 HOX-TFs and CTCF . To test this possibility , we made use of the proximity ligation assay ( PLA ) [26] . The PLA assay allowed us to assess protein-protein interactions in situ , in a quantifiable and sensitive manner . We expressed FLAG-tagged HOXA10 ( Group 1 ) in chicken DF1 cells and performed the PLA assay using αFLAG antibody and an endogenous αCTCF antibody . We readily detected CTCF-HOXA10 interaction in the nucleus that was almost as strong as the interaction of CTCF with RAD21 , which we used as a positive control ( Fig 4D ) . We also performed the same assay with CTCF and the Group 2 HOXD13 protein , for which our ChIP-seq data had predicted a weaker interaction . In this case , we measured a signal above our negative control ( DF1 cells expressing CTCF alone ) , but less than for the CTCF-HOXA10 interaction ( Fig 4D and 4E and S11 Fig ) .
In this study , we systematically compared the effect of nine limb-bud expressed HOX-TFs on the differentiation and gene regulation of primary mesenchymal limb bud cells . Hierarchical clustering of the regulated genes delineated two groups of HOX-TFs: HOX10/11/D9 , and HOXD12/HOXD13 that , during limb development , are expressed in the stylo/zeugopod and autopod , respectively . The distinction between these two groups is in accordance with genetic experiments in mice demonstrating that Hoxd12 , but not Hoxd11 is able to substitute for a loss of Hoxd13 [9] . Another interesting observation was that HOXA9 and HOXA13 clustered separately from the other factors . Differences between HOXA/D9 and HOXA/D13 paralogs , in contrast to the more similar HOXA/D10 and HOXA/D11 paralogs , were also apparent in their distinct effects on chMM differentiation . The differences between HOXA9 and HOXD9 ( or HOXA13 and HOXD13 ) might be attributed to the fact that the HOX9 and HOX13 paralog groups are the only posterior HOX-TFs which retained all four copies of the genes , thereby reducing the selective pressure on each paralog and allowing their neo-functionalization [3] . The expression analysis further revealed that overexpression of HOXA9 had a negative regulatory effect on most posterior HOX genes , especially its direct genomic neighbours in cis , HOXA10 and HOXA11 , further confirming the observation from hierarchical clustering that this factor is different from the other studied HOX genes . HOXA/D13 were not negatively regulated by HOXA9 but upregulated by all other HOX-TFs , demonstrating some degree of auto-regulation . Finally , although Hoxa13 and Hoxd13 are partially redundant in loss-of-function experiments [27] , we here find that they induce rather different regulatory programs and bind to different genomic regions despite having very similar biochemical DNA binding affinities . This highlights that their divergent molecular function could essentially result in similar developmental outcomes and their observed genetic redundancy reflects an additive effect of both TFs rather than true molecular redundancy . Our systematic comparison focused on the effects of individual HOX-TFs and their genome-wide binding . However , HOX-TFs are rarely expressed alone in vivo , but are rather co-expressed in overlapping patterns and exert their specific function in this biochemical context . Although HOX-TFs induced distinct effects in our experiments , their combinatorial or antagonistic action in vivo might play an important role in the developing embryo . Investigation of the in vitro sequence specificity of individual HOX-TFs showed that their homeodomains bind largely similar sequences [11 , 12] . Subsequent studies , however , revealed that the binding of cofactors changes the original HOX binding site resulting in recognition sites that are markedly different [10 , 16 , 17] . Both observations are reflected in the results of our ChIP-seq experiments . The low number of direct binding sites in HOX-TF peaks found in our experiments is in concordance with results from Drosophila , where low-affinity binding sites for the HOX-TF Ultrabithorax ( Ubx ) in complex with its cofactor Extradenticle ( Exd ) were shown to be biologically more significant [17] . Our analysis also highlights the role cofactors play in directing HOX-TF binding . The primary motif for both HOX11 paralogs was in many ways different from the in vitro determined monomer specificity and rather revealed a composite binding site like the one bound by a HOXA10-PBX4 dimer [10] . Furthermore , our data indicate a relationship between HOX-TFs and the AP1 class of TFs . AP1 binding sites were found in 5% of all HOX-TF peaks and JUN and FOS were also strongly upregulated by all HOX-TFs , suggesting a mechanism of cofactor cross-regulation . To our knowledge , AP1 has not been linked to limb patterning or HOX-TFs . However , these factors are known to be involved in a wide array of developmental and cell differentiation processes [28] and our results suggest AP1 may potentially have a role in mediating HOX-driven limb patterning . PCA analysis separated the HOX-TF binding sites in two subgroups along PC2 . We tried to identify the underlying cause for this distinction between HOX-TF binding sites and found co-binding with CTCF to correlate with Group 1 HOX-TF binding . We also describe CTCF as a novel cofactor of Group 1 HOX-TFs . CTCF/cohesin are now well-established factors with important functions in the spatial organization of the genome into topologically associating domains ( TADs ) [29–31] . Among other functions , CTCF/cohesin have been shown to directly mediate enhancer-promoter contacts [25 , 32 , 33] , and the co-occurrence of HOX-TF and CTCF binding sites therefore might result from the presence of CTCF at HOX-bound enhancers . However , the co-occupancy of CTCF/cohesin and HOX-TFs throughout the genome points to a possible role for this type of developmental TFs beyond enhancer-promoter communication . In contrast to our gain-of-function approach , loss of HOXA/D13 TFs has resulted in a miss-regulation of the HoxD locus during limb development , where two adjacent TADs regulate the gene expression in the proximal and distal limb , respectively [34 , 35] . Specifically , HOXA/D13 proteins did not regulate individual enhancers , but rather restructured the chromatin architecture of the locus in a way so that contacts with one ( the telomeric ) TAD were repressed , whereas contacts with the other ( centromeric ) TAD were promoted [35] . A related observation was recently reported in Drosophila for CTCF/Cohesin and Smad-TFs , which are the transcriptional effectors of TGFß/BMP signaling [36] . The Smad-TFs co-localized in a CTCF-dependent manner to CTCF binding sites within TADs and might be involved in sculpting the TAD to enable transcriptional regulation . The observed connection of certain developmental TFs with CTCF/cohesin architectural proteins suggests an important fundamental regulatory role for HOX and other TFs that extends beyond the control of individual gene expression .
HOX and CTCF coding sequences were amplified from chicken embryonic limb buds cDNA ( HH27 ) and cloned into RCASBP-viruses as previously described [18] . DF1 cells were transfected in a 6 cm dish with 3 μg of each RCASBP ( A ) plasmid using Polyethylenimine ( Polyscience Inc . #24765–2 ) and NaCl . Cells were expanded and stressed on starvation media whereupon the supernatant was harvested on three consecutive days . The supernatant was then centrifuged to produce the concentrated viral particles of high titer , 108 viral particles/ml or higher . The infection of chMM cultures and the histological assessment were performed as described elsewhere [18] . For the quantification of chondrogenesis , cultures after 3 , 6 , 9 , 12 and 15 days post-infection were fixed and stained with Alcian Blue . After 2 washes with 1 x PBS , the quantification of incorporated Alcian Blue was determined by extraction with 6 M guanidine hydrochloride , followed by photometric measurement at A 595 nm . The expression level of cloned constructs was controlled with Western Blot . Cells were lysed using RIPA buffer and immunoblotted using m-αFLAG M2 1:1 , 000 ( Sigma , F1804 ) ( S1D Fig ) . Chromatin Immunoprecipitation was performed as described previously [18] . Briefly , chMM cultures were harvested after 6 days of culture by adding digestion solution ( 0 . 1% collagenase ( Sigma , #C9891 ) and 0 . 1% Trypsin in 1x PBS ) to obtain a roughly single-cell suspension . Cells were taken up in 10 ml cold chMM ( DMEM: HAMF11 with 10% FBS , 10% CS , 1% L-glutamine and 1% Penicillin-Streptomycin ) medium and fixed for 10 min on ice with 1% formaldehyde . The extraction of nuclear lysate was performed as described in Lee , Johnstone ( 37 ) and chromatin was sonicated with a Diagenode Bioruptor ( 45 cycles—30-sec pulse , 30-sec pause , HI power ) . For ChIP , 25–35 μg of chromatin was incubated with 6–8 μg of antibody overnight . The next day blocked magnetic beads were added and incubated overnight , followed by 6 washes with RIPA and one with TE buffer [37] . After elution , the preparation of the library for pulled down DNA was performed as described previously [18] . Cells from harvested chMM cultures were separated prior to fixation of the ChIP samples and RNA was isolated from these cells using an RNaeasy Qiagen kit . RNA-seq libraries were constructed as described previously [18] , by selecting for fragment sizes between 300–500 bp and sequenced single-end 50 bp using Illumina technology . DF1 cells were transfected with RCASBP ( A ) -3x FLAG-HOXA10 , RCASBP ( A ) -3x FLAG-HOXD13 , or RCASBP ( B ) -HA-CTCF , respectively . The cells were cultured for at least 6 days to ensure a high cellular infection rate . Upon confluency cells were transferred to 10 mm cover slips and further incubated for one day . Cells were fixed for 10 min with 4% PFA , blocked with TSA ( 10% horse serum , 0 , 5% PerkinElmer blocking reagent [#FP1020] and 0 . 01% Triton-X-100 in 1x DPBS ) and incubated with appropriate primary antibodies ( in 10% horse serum in 1x DPBST ) overnight at 4°C . Primary antibody combinations were: 1 ) FLAG-HOX and CTCF interaction: m-αFLAG M2 and rb-αCTCF; and 2 ) HA-CTCF and RAD21 interaction: m-αHA and rb-αRAD21 . Antibody concentration for PLA were tested and used as follows: m-αFLAG M2 1:20 , 000 ( Sigma , F1804 ) , m-αHA . 11 1:8 , 000 ( BioLegends , #901501 ) , rb-αCTCF 1:20 , 000 ( ActiveMotif , #61311 ) and rb-αRAD21 1:1 , 000 ( Abcam , ab992 ) . After primary antibody incubation , the PLA assay was performed using the Duolink In Situ Fluorescence Kit ( Sigma , #DUO92101-1KT ) according to manufacturer’s instructions . Protein-protein interactions were analyzed by using confocal imaging on a Zeiss LSM700 and the Axiovert Zen software . For the quantification of PLA experiments , the contacts in several independent frames were counted using ImageJ and divided by the number of nuclei in the frame . The PLA experiments were performed in at least two independent experiments .
RNA-sequencing reads were mapped to the chicken reference genome galGal4 using the STAR mapper [44] ( splice junctions were based on RefSeq/ENSEMBL gene annotations; options included: alignIntronMin 20 , alignIntronMax 500000 , outFilterMultimapNmax 5 , outFilterMismatchNmax 10 , and—outFilterMismatchNoverLmax 0 . 1 ) . Read counts for individual genes were generated for a gene list combining the RefSeq ( galGal4 ) and ENSEMBL ( release 75 ) gene annotations . Log2 Fold changes for differential expression were calculated using DEseq2 [19] . The top 50 regulated genes were filtered according to p-value < 10−5 , a minimum base mean >30 and a fold change >2 . For hierarchical clustering , all genes were included that were among the top 50 regulated genes in at least one of the datasets . The log2-transformed fold changes , as compared with control cultures , were then used as the input for the R heatmap3 hierarchical clustering algorithm . The RCAS-virus is transcribed as a polycistronic mRNA that is spliced into three distinct isoforms , all containing the HOX-TF message and only one of which is spliced to code for the virally expressed tagged HOX-TF . To deconvolve the HOX-coding splice isoforms from all viral mRNA , we counted the number of RNAseq reads that could be uniquely associated to each of the three possible isoforms and then calculated the fraction of HOX-specific splice variants ( SD-SA2 ) . We then multiplied the HOX-TF RPKM with this ratio to obtain the effective RPKM and compared it to the HOX-RPKMs from HH23 wing buds , where all nine relevant HOX-TFs are at least partially expressed [45] . In order to generate a correct estimate of the expression values for the HOXA genes we had to manually alter the chicken Galgal4 annotation . In the Galgal4 annotation ( but not in Galgal3 or Galgal5 ) the HOXA cluster and its genes are annotated in triplicate . We manually changed the annotation so that every additional copy of the gene would be considered a new gene . We then mapped the RNA-seq samples to the altered annotation and used it only to assess the overexpression levels and the HOX autoregulation . | Hox genes encode transcription factors that determine the vertebrate body plan and pattern structures and organs in the developing embryo . Despite decades of effort and research on Hox genes , little is known about the HOX-DNA binding properties in vivo . This lack of knowledge is mainly due to the absence of appropriate antibodies to distinguish between different HOX transcription factors . Here , we adapt a cell culture system that allows us to investigate HOX-DNA binding on a genome-wide scale . With this approach , we define and directly compare the genome-wide binding sites of nine posterior HOXA and HOXD transcription factors . We report that the in vivo HOX binding specificity differs from the in vitro specificity and find that HOX-TFs largely rely on co-factor binding and not only on direct HOX-DNA binding . Finally , we identify a novel HOX co-factor , a genome architecture protein , CTCF suggesting a possible interplay between HOX-TF function and chromatin architecture . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Bioinformatic",
"Analyses"
] | [
"livestock",
"chemical",
"characterization",
"gene",
"regulation",
"regulatory",
"proteins",
"enzymology",
"vertebrates",
"dna-binding",
"proteins",
"animals",
"transcription",
"factors",
"sequence",
"motif",
"analysis",
"enzyme",
"chemistry",
"research",
"and",
"analysis",... | 2017 | Genome-Wide Binding of Posterior HOXA/D Transcription Factors Reveals Subgrouping and Association with CTCF |
Telomere shortening is associated with aging and age-associated diseases . Additionally , telomere dysfunction resulting from telomerase gene mutation can lead to premature aging , such as apparent skin atrophy and hair loss . However , the molecular signaling linking telomere dysfunction to skin atrophy remains elusive . Here we show that dysfunctional telomere disrupts BMP/pSmad/P63 signaling , impairing epidermal stem cell specification and differentiation of skin and hair follicles . We find that telomere shortening mediated by Terc loss up-regulates Follistatin ( Fst ) , inhibiting pSmad signaling and down-regulating P63 and epidermal keratins in an ESC differentiation model as well as in adult development of telomere-shortened mice . Mechanistically , short telomeres disrupt PRC2/H3K27me3-mediated repression of Fst . Our findings reveal that skin atrophy due to telomere dysfunction is caused by a previously unappreciated link with Fst and BMP signaling that could be explored in the development of therapies .
Telomeres consist of ( TTAGGG ) n DNA repeats and associated proteins that locate at chromosome ends , maintaining chromosomal stability and cell proliferation . The telomerase complex consists of a telomerase RNA component ( TERC ) and the reverse transcriptase catalytic subunit ( TERT ) , and adds telomere repeats to chromosome ends to offset the loss of telomere sequences that occurs due to the end-replication problem , the inability of DNA polymerase to replicate fully the lagging DNA strand [1] . In the absence of telomerase , telomeres shorten progressively with cell division , ultimately leading to loss of telomere protection and a DNA damage response that induces senescence or cell death . Telomere shortening is closely tied to organism aging and premature aging and associated diseases [2–5] . Skin atrophy and hair loss are general phenomena associated with age [6] . Moreover , patients with the mutation of telomerase components ( e . g . Dyskerin , TERT , TERC ) exhibit telomere shortening and skin atrophy [7] . It has been shown that short telomeres impair differentiation and development of the epidermis , and cause skin atrophy and loss of hair follicles , in association with epidermal stem cell dysfunction with aging [8–10] . However , the molecular signaling underlying short telomeres-associated skin atrophy or degeneration and hair follicle loss remains elusive . Embryonic stem ( ES ) cells are able to spontaneously differentiate into three embryonic germ layers ectoderm , mesoderm , and endoderm by standard test of embryoid body ( EB ) formation . This method has been extensively used to investigate signaling pathways that control ES cell differentiation towards various cell lineages [11–13] , including epidermis [14–16] . Telomere length is critical for developmental pluripotency and differentiation capacity of ES cells or iPS cells [17–20] . We attempted to investigate how short telomere compromises epidermal lineage specification and differentiation initially by using ES cell lines with different telomere lengths , derived from Terc knockout ( Terc–/– ) mice [17] . We showed that telomere lengths affected differentiation of ES cells into epidermis . We further validated that short telomeres impeded epidermal differentiation in the adult telomerase-deficient , telomere shortened mice . Moreover , we investigated potential regulatory mechanisms of telomere length on epidermis differentiation .
To investigate the differentiation defects associated with short telomeres , we initially performed in vitro differentiation experiments by standard EB formation test using mouse ES cells with various telomere lengths due to telomerase ( Terc–/– ) deficiency ( Fig 1A and 1B , S1A Fig ) . Telomeres were longest in wild-type ( WT ) ES cells , shorter in heterozygous ( Terc+/– ) and early generation ( G1 ) Terc–/–ES cells , and critically short or lost in late generation ( G3 and G4 ) Terc–/–ES cells ( Fig 1C and 1D ) , as we previously reported [17] . Late generation ( G3 and G4 ) Terc–/–cells also exhibited short telomeres by day15 of differentiation ( Fig 1C and 1D ) . Upon differentiation , WT ES cells showed significantly reduced expression of pluripotency marker genes such as Oct4 and Nanog ( S1B and S1C Fig ) . However , G3 and G4 Terc–/–ES cells maintained expression of Nanog and Oct4 at relatively high levels , and low methylation at Nanog promoter ( S1D Fig ) , consistent with the finding using Tert–/–ES cells also with critically short telomeres [18] . Expression levels of genes related to endoderm , mesoderm and neuro-ectoderm did not differ between WT ES cells and ES cells with short telomeres , suggesting that shortening of telomeres does not significantly affect the differentiation of these germ layers ( Fig 1E and 1G ) . Notably , expression levels of genes important for epidermal ectoderm differentiation were consistently reduced in telomere shortened ( G3/G4 Terc–/– ) ES cells following differentiation as compared to WT ES cells ( S1E and S1F Fig ) . During mouse embryo development , epidermal progenitors are specified at around embryonic day 8–12 ( E8-12 ) , later than that of neural induction . Expression of Keratin 14 ( K14 ) was at very low level on day 8 of differentiation ( Day 8 ) and was sharply increased on day 15 in WT cells . K14 was also low on day 8 in G3/G4 Terc–/–cells , but dramatically reduced on day 15 , as compared to WT cells . Consistently , expression levels of K5 ( epidermal basal cell marker ) , K1 ( epidermis marker of skin ) and K4 ( epidermis marker in stratified epithelia ) [21] , in the differentiated G3/G4 Terc–/–cells were also significantly lower than that in WT cells ( S1F Fig ) . p63 as one of the earliest genes for epidermal lineage is expressed as early as E7 . 5 , identifies epidermal keratinocyte stem cells , and is required for epidermal differentiation [22–24] . p63 also is expressed earlier than does K14 during differentiation of human ES cells into keratinocytes [15] . Consistently , p63 expression was detectable in WT , Terc+/– , and G1 Terc–/–cells by day 7–8 of differentiation , earlier than that of Keratins , but only minimal in G3/G4 Terc–/–cells . p63 level was further increased by day 15 in WT , Terc+/– , and G1 Terc–/–cells , but much lower in G3 and G4 Terc–/–cells ( S1E Fig ) . Consistent with the qPCR data , protein levels of both K14 and P63 at day 15 were also greatly reduced in G3/ G4 Terc–/–cells as compared to WT cells ( Fig 1E ) . Immunofluorescence microscopy showed specific staining of P63 in the nuclei and K14 in cytoplasm and membrane in WT cells but much reduced staining in some G4 Terc–/–cells ( Fig 1F ) . These data indicated that short telomeres lead to decreased expression of P63 and K14 and that telomere-shortened stem cells may fail to stratify in the differentiation into epidermal lineage . To examine the impacts of short telomeres on the differentiation capacity in vivo , we initially performed standard teratoma formation test [12 , 25] . Both WT and G4 Terc–/–ES cells were able to differentiate into three germ layers , including endoderm , mesoderm , and neural ectoderm revealed by histology ( Fig 2A ) . However , epidermis lineage was reduced in teratomas differentiated from ES cells with short telomeres , in contrast to that of WT ES cells ( Fig 2A and 2B , S2A Fig ) . Structures in size or number identified by epidermis marker K14 or epidermal stem cell marker P63 were reduced in the sections of teratomas from G4 Terc-/- ES cells , compared with those from WT ES cells ( Fig 2B ) . Relative mRNA levels of p63 and K14 in teratomas derived from G4 Terc–/–ES cells also were lower than those from WT ES cells ( Fig 2C ) . Similar phenotypes also can be observed in the adult G3 Terc–/–mouse skin . Epidermis marked by co-immunostaining of P63 and K14 and by histology was thinner on average in skin of two-three month old G3 Terc–/–mice , compared with age-matched WT mice ( Fig 2F and 2G , S2B and S2C Fig ) . Additionally , hair follicles were readily seen in dermis of WT mice but fewer in G3 Terc–/–mice ( only 50% of WT mice ) ( Fig 2D and 2E , S3B Fig ) . Number of hair follicles was calculated based on at least 10 fields of view under microscopy . In WT mouse skin , the hair follicles are structurally intact with an average of 4 to 5 per field of view . However , hair follicles drop sharply in their numbers and loses the typical structure in the G3 Terc–/–mouse skin ( Fig 2D and 2E ) . Both in vitro and in vivo results validated that short telomeres reduce epidermal commitment . To understand the mechanisms underlying short telomeres-affecting ES cell differentiation towards epidermal lineage , we performed microarray analysis of G4 Terc–/–ES cells compared with WT ES cells . Interestingly , Follistatin ( Fst ) , a negative regulator of Smad pathway which is critical in epidermis commitment , was expressed at higher level in G4 Terc–/–than in WT ES cells at day 0 ( Fig 3A ) . qPCR analysis validated that expression levels of Fst were higher in G3 and G4 Terc–/–than in WT ES cells ( Fig 3B ) . Western blot also confirmed that Fst protein level was indeed higher in G4 Terc–/–ES cells than in WT ES cells during differentiation ( Fig 3C , left panel ) . Given that Fst is a secreted protein , we also examined Fst protein levels in the culture media for both cell lines . Fst protein was highly abundant in the culture media of G4 Terc–/–ES cells , but barely detectable in that of WT ES cells ( Fig 3C , right panel ) . Furthermore , robust cytoplasmic and membrane staining of K14 and nuclear P63 were observed at day 15 in differentiated WT cells , but their expression levels were markedly reduced in differentiated G4 Terc–/–cells , where higher Fst fluorescence signals with dotted staining still were readily visible in the cytoplasm of or around the differentiated cells , compared with lower Fst fluorescence in WT cells ( Fig 3D ) . Additionally , Fst protein level was higher in G4 Terc–/–teratomas than in control teratomas ( WT and Terc+/– ) ( Fig 3E ) . Compared to WT teratomas , G4 Terc–/–teratomas exhibited strong Fst immunofluorescence spotted inside or outside the cells , coincided with less and weak fluorescence staining of K14 and P63 ( Fig 3F and 3G ) . These data provide further evidence that higher expression level of Fst is linked to short telomere . Similar results were obtained from the skin of adult mice . Two-three month old G3 Terc–/–mice displayed thinner epidermis and skin atrophy compared to the age-matched WT mice , consistent with previous studies [8 , 26] . K14 level also was reduced in the epidermis of G3 Terc–/–adult mice , accompanied by increased expression of spotty Fst as compared to WT mouse epidermis ( S3A Fig ) . Immunofluorescence staining of Fst in teratomas or tissues appears to be dotted in pattern , somewhat different from the immunostaining in cultured cells , probably because Fst can be locally confined with specific structure in tissues , whereas it diffuses in and around the cultured cells . In addition , G3 Terc–/–mice displayed defective hair follicle development as evidenced by notably reduced number of hair follicles , with reduced expression of K14 and increased Fst , as well as impaired bulb and bulge at the basal follicles where progenitor cells reside , in contrast to the intact bulb ( hair germ ) and bulge in WT mice ( S3B Fig ) . Taken together , short telomeres lead to excessive expression of Fst , which is incompatible with epidermal stem cell specification and stratification of skin and hair follicles . Collectively , these findings suggested that short telomere specifically prevents the transition from the common ectodermal progenitor state into the epidermis fate . Bone morphogenesis protein 4 ( BMP4 ) signaling is known to be activated in the embryo at the time of ectodermal fate determination , inhibits premature neural differentiation while inducing epidermis development , and can act through phosphorylation and nuclear accumulation of Smad1/5/8 [27–30] . Differentiation of epidermal cells appears to be controlled , in part , by BMP4 [31] . Fst is an antagonist of BMP4 . We next asked if the Fst-BMP-Smad1/5/8 signaling pathway plays a critical role in epidermal differentiation . In the wild-type ES cells , the up-regulation of BMP4 , BMP7 , Smad1 , and the down-steam target ( Gata1 ) during differentiation indicates that this pathway plays an important role in normal differentiation and development of epidermis ( S4 Fig ) . Compared with those of WT , Terc+/– , or G1 Terc–/–ES cells , levels of phosphorylated Smad1/5/8 were reduced in G3/G4 Terc–/–ES cells during differentiation ( Fig 3H ) , suggesting that this pathway is suppressed by short telomeres . Gene expression upstream of this pathway seemed to be not affected by short telomeres . Downstream target genes of this signaling pathway such as Gata1 were expressed at lower levels in G4 Terc–/–than in WT cells by day 8 and day 15 of differentiation ( S4 Fig ) . These data suggested that short telomeres suppress BMP/pSmad signaling following differentiation . Above data imply that elevated expression of Fst resulting from short telomere might lead to reduction of pSmad1/5/8 , P63 and K14 , and thus defective epidermal stem cell specification and differentiation . To further validate this concept , we generated Fst overexpression ( OE ) ES cell line ( Fig 4A ) and performed EB differentiation test using WT ES cell line as control . Western blot showed that Fst OE ES cells expressed p63 and K14 at reduced levels on day 8 and day 15 of differentiation , which was also confirmed by immunofluorescence microscopy ( Fig 4B and 4C ) . Notably , in the differentiated Fst OE cell culture , areas with intensive Fst fluorescence indicative of high expression level exhibited minimal K14 staining , and yet areas with low Fst fluorescence displayed strong K14 or p63 staining ( Fig 4C ) . Hence , high Fst level is discordant with expression of P63 and K14 . Consistently , pSmad1/5/8 was decreased in Fst OE cells ( Fig 4B ) . These data suggest a conserved but new role of Fst in negatively regulating pSmad-signaling pathway during epidermal ectoderm induction . To test whether reducing Fst can de-repress down-stream genes/signaling for epidermis , we knocked down Fst by RNA interference in the differentiated G4 Terc–/–cells . Effective knockdown of Fst by shRNA in differentiated G4 Terc–/–cells up-regulated the levels of pSmad1/5/8 and P63 ( Fig 4D and 4E ) . Fst downgregulation by RNA interference in the differentiated G4 Terc–/–cells rescued P63 but not fully rescued K14 expression . This may be explained by three potential reasons . Changes in the expression level of K14 could be delayed following P63 expression during epidermal differentiation . Factors other than Fst alone also might be involved in regulation of K14 expression . Alternatively , the regulation of Fst-P63-K14 may slightly differ in differentiated ES cells compared with undifferentiated ES cells as model . Nevertheless , these data further support the notion that excessive expression levels of Fst negatively regulate pSmad1/5/8 signaling and p63 , weakening epidermal stem cell specification and differentiation . Then , we tested whether rejuvenating telomeres in ES cells with short telomeres can repress Fst . Using CRISPR/Cas9 technology , we successfully knocked in Terc in G4 Terc–/–ES cells and obtained several Terc-repaired ES cell lines ( two lines are shown in Fig 5A ) . These Terc repaired ( TR ) ES cell lines exhibited much longer telomeres than did their parental G4 Terc–/–ES cell line after culture for 10 passages . Yet , their telomeres were still shorter than those of WT cells as revealed by qPCR and QFISH ( Fig 5B and 5C ) , presumably because of inadequate passages , even though the telomerase activity was recovered ( Fig 5D ) . Frequency of telomere loss was significantly reduced in Terc repaired ES cell lines , in contrast to that of G4 Terc–/–ES cells ( Fig 5C ) . We repeated the in vitro differentiation assay with WT , G4 Terc–/– , and Terc repaired ES cell lines . On day 15 of differentiation , telomere length was also rescued in Terc repaired cells compared with Terc–/–cells ( Fig 5B ) . Terc repaired cells showed reduced level of Fst and noticeably increased protein levels of P63 and K14 as compared to those of G4 Terc–/–cells ( Fig 5E ) , which were confirmed by immunofluorescence microscopy ( Fig 5F and 5G ) . These results suggested that epidermal differentiation could be rescued by repairing Terc and restoration of telomere length . The critical question was how short telomeres result in excessive Fst expression . Fst gene is located at the subtelomere region of the long arm of chromosome 13 , whose expression might be regulated by telomere position effect ( TPE ) [32] . To reveal the telomere state of chromosome 13 , we performed immunofluorescence microscopy to detect the chromosome 13 using the chromosome specific probe followed by telomere FISH . Notably , one pair of chromosome 13 in G4 Terc–/–ES cells constantly displayed telomere signal-free ends , indicative of telomere loss , in contrast to four intact telomere signals of WT ES cells ( Fig 6A ) . Moreover , chromosome fusion or translocation was found in the chromosome 13 with loss of telomere signals in G4 Terc–/–ES cells ( Fig 6A ) . Loss of telomeric repeats leads to a change in the heterochromatic architecture with decreased H3K9me3 abundance at telomeres/subtelomeres [33] . We tested whether epigenetic modifications are implicated in regulation of Fst . Both DNA methyltransferases Dnmt3a and 3b were expressed at lower levels in G4 Terc–/–ES cells than in WT cells , but Dnmt3b expressed at higher levels following differentiation ( S5A Fig ) . By ChIP-qPCR analysis using specific primers and Dnmt3b antibody , levels of Dnmt3b at Fst loci did not differ between G4 Terc–/–and WT ES cells ( S5B Fig ) . Also , Fst promoter showed only low methylation levels in G4 Terc–/–and WT ES cells like that of MEF ( S5C Fig ) . Methylation levels at subtelomeres of chromosome 13 were greatly reduced in G4 Terc–/–ES cells , but markedly increased in G4 Terc–/–cells following differentiation , compared with WT ES cells ( S5D Fig ) . These data suggest that Fst promoter methylation may not directly contribute to excessive Fst expression due to short telomere . We analyzed the abundance of histone modifications of H3K4me3 , H3K9me3 and H3K27me3 by western blot . H3K4me3 abundance seemed not to differ between G4 Terc–/–and WT ES cells , while H3K9me3 and H3K27me3 abundance were slightly reduced in G3/G4 Terc–/–ES cells , compared with WT , heterozygous , or G1 Terc–/–ES cells ( S6A Fig ) . Also , G4 Terc–/–ES cells exhibited decreased H3K9me3 immunofluorescence and foci at heterochromatin and telomeres prior to and after differentiation , compared with WT ES cells ( S6B Fig ) . Furthermore , we performed ChIP-qPCR analysis to examine the abundance of H3K9me3/2 , H3K9ac and H3K27me3 at Fst promoter loci using β-actin as a control . Enrichment of H3K9me3 , H3K9Ac , and H3K9me2 at Fst promoter was low and showed no significant difference between WT and G4 Terc–/–ES cells ( Fig 6B ) . However , H3K27me3 was highly enriched at Fst promoter . Importantly , H3K27me3 level was markedly reduced at all five loci of Fst promoter in G4 Terc–/– , compared with that of WT ES cells ( Fig 6B ) . By luciferase reporter assay , the Fst promoter activity was higher in G4 Terc–/–than in WT ES cells ( S7A Fig ) . We further examined expression levels of Eed , Suz12 , Ezh1 and Ezh2 which are catalytic components of Polycomb repressive complex PRC2 and potentially tri-methylate H3K27 to repress gene expression and that are shown to play important roles in skin stem cell function and differentiation [34 , 35] . Expression levels of Ezh1 and Ezh2 are reduced in G4 Terc–/–ES cells as compared to WT ES cells ( S7B Fig ) . Telomere-repaired ES cells partially restored Ezh1/2 expression , together with increased H3K27me3 enrichment at Fst promoter ( S7C and S7D Fig ) . Pluripotent marker genes Nanog and Oct4 were also down-regulated during differentiation of Terc repaired G4 ES cells , like those of WT ES cells ( S7E and S7F Fig ) . It is interesting to note that Ezh2 expression level in one Terc-repaired ES cell line ( A49 TR3 ) was not recovered well , and coincidently this clone had relatively shorter telomere than that of WT ES cells ( Fig 5C ) . These results further suggest that short telomeres reduce H3K27me3 enrichment at Fst promoter , likely together with reduced Ezh1 and Ezh2 levels , de-repress Fst , and these together may contribute to excessive expression of Fst . Excessive Fst further down-regulates p63/K14 through disrupting BMP4/pSmad signaling ( Fig 6C ) .
Based on the data obtained from both ES cell differentiation in vitro and in vivo , we propose that functional telomere is important for suppressing Fst to prevent its overexpression and to maintain normal expression of P63 and K14 during epidermal stem cell specification and differentiation . Short telomere disrupts PRC2- H3K27me3-mediated repression of Fst , which leads to excessive Fst expression . Consequently , excessive Fst suppresses BMP/pSmad signaling , reducing P63 and keratins and resulting in epidermal differentiation defects and skin atrophy . This model links dysfunctional telomeres to skin atrophy and hair follicle loss by disrupting Fst/BMP/pSmad/P63/K14 signaling . This study also provides additional evidence in supporting that ES cell differentiation model is a powerful alternative tool to discover novel signaling and mechanisms that are involved in in vivo cell lineage specification at very early developmental stages that might not be readily revealed in live mouse model and particularly in humans [11] . The differentiation assay used in our study shows that the dynamics of P63 and K14 in mouse ES cell is similar to that of human ES cells and mouse embryonic skin development [14] . Our results also confirmed that P63 is a master regulator for K14 , K5 and other epidermal genes [36 , 37] , and that BMP4/pSmad signaling pathway can activate P63 [38] . BMP4 negatively regulates neural induction and promotes epidermogenesis during differentiation of mouse ES cells [31] , whereas blocking BMP signaling facilitates differentiation of human ES cells into neural lineages [28] . Mounting evidence supports the notion that telomere dysfunction is accompanied by symptoms of abnormal epidermis [10 , 39–42] . Mice with critically short telomeres exhibit symptoms , including epidermal abnormalities such as poor wound healing , ulcerative skin lesions , early hair loss and early hair graying [2 , 8 , 10 , 43] . We show that short telomeres lead to reduced expression of P63 and declined epidermal stratification and formation , linking to skin atrophy . Study of P63-null mice demonstrates important roles of P63 in orchestrating first epidermal stratification [44 , 45] . p63-null mice exhibit striking defects in embryonic epidermal morphogenesis [46] , and also suffer from diminished stem cell renewal capacity [47] . Moreover , TAp63 serves to maintain adult skin stem cells and prevents premature tissue aging [45 , 48 , 49] . Hence , P63 is required to maintain epidermal stem cell renewal while allowing K14 expression and epidermal differentiation [24] . Short telomeres cause stem cell failure [50] , and also impair the ability of epidermal stem cells to mobilize out of the hair follicle niche , and thus skin and hair growth [26] . On the other hand , hyper-long telomeres are advantageous for skin regeneration compared with normal length telomeres [51] . Our data provide novel molecular mechanisms of linking short telomeres to reduced pSmad signaling and P63 and thus declined epidermal differentiation . Moreover , excessive Fst expression resulting from short telomere negatively regulates BMP/pSmad/P63 pathways in the epidermal stem cell specification and differentiation . It has been reported that Fst is an antagonist of BMPs by blocking binding of BMP with its receptor [52] . Excessive Fst may compete with BMPs and inhibit BMP-pSmad signaling . Telomere re-elongation successfully achieved by CRISPR/Cas9-mediated knock-in of Terc represses Fst and recovers expression of P63 and K14 . Consistently , telomerase reintroduction into mice with critically short telomeres is sufficient to elongate telomeres in skin keratinocytes and to correct epidermal hair follicle stem cell defects , and rescues skin and hair growth defects [26] . These data also may explain the early findings that Fst-knockout mice die within hours of birth but show thicker epidermis [53] . Likewise , deletion of Fst results in enhanced keratinocyte proliferation in the tail epidermis of these animals and an earlier onset of keratinocyte hyperproliferation at the wound edge after skin injury , suggesting that Fst regulates epidermal homeostasis and also wound repair [54] . In agreement , Fst-overexpression transgenic mice are characterized by a thinner dermis and epidermis , reduced density of the dermis and smaller hair follicles , indicative of skin atrophy , and a severe delay in wound healing observed after injury [55] . Moreover , mice that overexpress Fst are smaller compared with their control littermates , and their body weight is significantly reduced . This phenotype is similar to that of late generation Terc–/–mice [2 , 56] ( also shown in this study ) , and these mice also exhibit severely impaired wound healing [2] . Coincidently , p63−/− mice have an impaired wound-healing response as well [48] . Together , these data support the idea that abnormal Fst/p63 signaling is implicated in short telomeres-associated skin atrophy and wound healing . Excitingly , mouse ES cells with hyper-long telomeres generate healthier chimera mice that also have longer telomeres and exhibit delayed aging and high capacity for skin wound healing [51] . Short telomeres can change expression of many genes and signaling pathways particularly with cell differentiation , as shown by the transcription profile data . We identified unique alterations of down-stream genes under regulation by the major TGFβ superfamily during differentiation of ES cells into epidermal lineage . Fst happens to be an evident negative regulator upstream in this pathway , and is up-regulated when telomere is short . We searched for mechanisms underlying short telomere-induced activation of Fst , and tested the hypothesis that repressive histone modification or DNA methylation may underlie telomere suppression of Fst . By ChIP-qPCR assay with selective related antibodies , we show that PRC2-mediated repression involving Ezh1/2 and H3K27me3 makes a major contribution to suppressing Fst . In fact , the regulatory region of Fst gene is characteristic of bivalent genes whose promoters are enriched for both activating mark by H3K4me3 and repressing mark by H3K27me3 and Ezh2 , primed for differential expression upon differentiation [57–59] . In mice , PRC2 has been found to be enriched in the progenitor cells of developing epidermis , regulates epidermal specification in mouse embryos and maintains hair follicle homeostasis [60 , 61] . H3K27me3 marks are enriched in a subset of epidermal differentiation gene promoters in undifferentiated cells and disappear on a subset of epidermal gene promoters upon differentiation [62] . Moreover , Ezh1 and Ezh2 repress premature differentiation and H3K27me3 is involved in early lineage specification of embryonic epidermis differentiation [60] . Interestingly , Ezh1/2 null skin progenitors show reduced H3K27me3 abundance and significant up-regulation of Fst [61] . Identification of hair follicle stem cell signature genes showed that Fst also is one of genes involving transit-amplifying ( TA ) progeny repressed by H3K27me3 , whereas BMP4 signaling is activated during this process likely induced by epigenetic shift to control by H3K4me3 and H3K79me2 [63] . Consistently , short telomere reduces H3K27me3 enrichment at Fst promoter , which leads to elevated Fst expression and defective epidermal specification and differentiation . Terc-repaired G4 Terc–/–ES cells rejuvenate telomere length to various degrees and partly restore H3K27me3-mediated suppression of Fst . Nanog and Oct4 are down-regulated following differentiation of Terc-repaired G4 Terc–/–ES cells like WT cells . Coincidentally , Tert-/- ES cells also have critically short telomeres and disrupted PRC2 function and low H3K27me3 enrichment at Nanog promoter , leading to defective suppression of Nanog during differentiation [18] . Fst-BMP4 signal pathway is known as a critical regulator for epidermal differentiation initiation and induced expression of p63 , which may coordinate with BMP4 to accelerate epidermal specification by regulating accumulation of H3K27me3 [64] . Deletion of p63 resulted in a significant decrease in signal of H3K27me3 mark [64] . We show that short telomeres can up-regulate Fst via reducing H3K27me3 at Fst promoter and decrease pSmad , resulting in declined expression of p63 . These findings suggest a complex feedback mechanism between H3K27me3 and Fst-BMP4-P63 . Fst , BMP4 , P63 , and H3K27me3 are key players in the orchestra that regulates epidermal differentiation . Another interesting phenomenon for Fst promoter is its hypomethylated state . Additionally , methylation levels at subtelomeres of chromosome 13 where Fst gene is located are also drastically reduced in G4 Terc–/–ES cells , compared with those of WT ES cells . Coincidently , Fst and other subtelomeric genes such as Tcstv1/3 in chromosome 13 are expressed at higher levels in G4 Terc–/–ES cells , but down-regulated in Terc-repaired G4 Terc–/–ES cells like WT cells ( S8 Fig ) . DNA hypomethylation could lead to decreased levels of H3K27me3 in ordinarily unmethylated regions [18 , 65] . Our data suggests that H3K9me3-mediated gene silencing does not play a direct role in repressing Fst . We indeed find a global reduction of H3K9me3 in G4 Terc–/–ES cells in which H3K9me3 also shows reduced co-localization with telomeres . These data suggest that telomere shortening-induced reduction of H3K9me3 at telomeres/subtelomeres may have a general impact on gene de-repression instead of a direct impact on Fst gene . Taken together , short or loss of telomere disrupts PRC2 function involving H3K27me3 and de-represses Fst . Elevated Fst inhibits pSmad/P63 signaling , leading to defective epidermal stem cell specification , stratification and differentiation . Rejuvenating telomere length can rescue these defects . We do not exclude the possibility that additional signaling pathways may also be involved in and/or cooperate with aberrant Fst/pSmad/P63 signaling in defective epidermal differentiation resulting from telomere dysfunction . Targeting Fst/pSmad/P63 pathway may have implications in ameliorating skin and hair degeneration associated with aging and telomere shortening .
All animal experiments were approved by the Institutional Animal Care and Use Committee at Nankai University ( License number 20140006 ) . All animal studies were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of Nankai University . All efforts were made to minimize the number of animals used by the experimental design . Two-three month old Terc deficient ( Terc–/– ) mice and wild-type mice in C57Bl/6 background , and immunodeficient mice were used in this study . Mice were housed and cared for in a pathogen-free facility at Nankai University . Terc–/–ES cells were generated from Terc deficient mice and cultured as previously described [17] . N33 ES cell line was derived from wild-type mice , heterozygous ( H1 ) ES cells from Terc+/–mice , and F19 , F35 , and A49 ES cell lines from G1 , G3 , G4 Terc–/–mice , respectively . These ES cells were maintained on mitomycin-C treated mouse embryonic fibroblasts as feeders in ES cell culture medium containing knockout Dulbecco’s modified Eagle medium ( KO-DMEM ) ( Invitrogen ) added with 20% fetal bovine serum ( Hyclone ) , 1000 U/ml LIF , 0 . 1 mM β-mercaptoethanol , 1 mM L-glutamine , 0 . 1 mM non-essential amino acids , 100 units/ml penicillin and 100 μg/ml streptomycin . pSpCas9 ( BB ) -2A-Puro ( PX459 ) was a gift from Feng Zhang ( plasmid # 48139 , Addgene ) . Guide RNAs were designed using the online design tool available at http://crispr . genome-engineering . org/ . PX459 was digested with BbsI and then gel purified . Two pairs of oligos including targeting sequences were annealed and cloned into the BbsI-digested PX459 vector . The Terc donor sequences were obtained based on mouse genomic sequence and the information provided in the original paper [43] . The Terc donor vector contained Terc flanked by 5’ ( Left ) and 3’ ( Right ) homology arms . The DNA fragments are individually amplified by proper primers and then cloned into the vector with proper enzymes . G4 Terc–/–ES cell line A49 was transfected with two PX459 and Terc donor plasmids using lipofectamine 2000 transfection reagent ( Invitrogen ) . Twenty-four hours later , 2 μg/ml puromycin was added into the culture medium for 7 days , clones were picked and the genomic DNA was extracted . PCR was performed with several pairs of primers to detect and obtain the genomic knock-in Terc repaired cell lines . ES cells are allowed to aggregate and form three-dimensional colonies known as embryoid bodies ( EBs ) [66] . Differentiation of ES cells was accomplished in a two-step process: ( 1 ) Embryoid body ( EB ) formation was obtained by using cell suspension and hanging drop method . Undifferentiated ES cells were trypsinized to obtain a single cell suspension , and EBs were formed in ES cell culture medium without LIF , in a definite number of cells in "hanging drops" for 4 days . ( 2 ) Then , EB were transferred to 24-well microwell plates with one EB per well . Daily microscopic observations were conducted to detect beating EBs . 10~15 EBs were transferred to 6-well microwell plates per well for protein , RNA , and DNA sample collection . Approximately 2×106 ES cells with different telomere length were injected subcutaneously into dorsal flanks of immunodeficient mice . Four weeks after the injection , the mice were humanely sacrificed and the teratomas were surgically dissected from the mice . Samples were weighed , fixed in PBS containing 3 . 7% formaldehyde , and embedded in paraffin . Sections were stained with hematoxylin and eosin for histological examination . The total RNA was isolated from samples using TriZol ( Invitrogen ) or RNeasy mini kit ( Qiagen ) according to the manufacturer's protocol . The purity and concentration of RNA were checked using Nanodrop technology ( Agilent ) . 2μg RNA was subjected to cDNA synthesis using M-MLV Reverse Transcriptase ( Invitrogen ) . Quantitative real-time PCR reactions were set up in duplicate with the FastStart Universal SYBR Green Master ( ROX ) ( Roche ) and run on the iCycler iQ5 2 . 0 Standard Edition Optical System ( Bio-Rad ) . Each sample was repeated at least twice and analyzed with Gapdh served as the internal control . Quantification of gene expression was based on the Ct ( Cycle threshold ) value . Melting curve analysis and electrophoresis were performed to control PCR products specificities and exclude nonspecific amplification . PCR Primers , designed using Primer5 and Gene Runner software , are listed in S1 Table . Cells were collected and washed with cold phosphate buffered saline ( PBS ) , then resuspended in cell lysis buffer containing 50 mM Tris ( pH 7 . 4 ) , 150 mM NaCl , 1 mM EDTA , 1 mM EGTA , 1 mM NaF , 20 mM Na4P2O7 , 1 mM Na3VO4 , 1%Triton X-100 , 10% glycerol , 0 . 25% deoxycholate and 0 . 1% SDS . 20 μg of proteins were separated on 10% SDS-polyacrylamide gels and transferred to polyvinylidene difluoride ( PVDF , Millipore ) membrane . Nonspecific binding was blocked by incubation in 5% nonfat dry milk in TBST at room temperature . Blots were then probed overnight at 4°C with primary antibodies against K14 ( ab7800 , Abcam ) , P63 ( ab124762 , Abcam ) , H3 ( ab1791 , Abcam ) , H3K4me3 ( ab1012 , Abcam ) , H3K9me3 ( 07–442 , Millipore ) , H3K27me3 ( 07–449 , Millipore ) , Smad1 ( #9743 , CST ) , pSmad1/5/8 ( #9511 , CST ) , pSmad2/3 ( #8828 , CST ) , Smad2/3 ( #5678 , CST ) , Fst ( ab64490 , Abcam ) , Dnmt3a ( ab13888 , Abcam ) , Dnmt3b ( ab13604 , Abcam ) , or β-actin ( sc1616R , Santa Cruz ) , washed and incubated for 2 h with secondary antibodies HRP conjugated donkey anti-Rabbit IgG ( NA934v GE Healthcare ) or goat anti-mouse IgG ( H+L ) ( ZB2305 ) . Protein bands were detected using ECL western blotting detection reagents ( WBKLS0100 Millipore ) . The band intensity was measured by software ImageJ and normalized to the intensity of β-actin . The relative expression level was calculated from the results of at least three independent experiments or samples and presented as mean ± SEM [67] . Cells were washed in PBS and stored at -20°C until subsequent DNA extraction . Genome DNA was prepared using DNeasy Blood & Tissue Kit ( Qiagen , Valencia , CA ) . Average telomere length was measured from total genomic DNA using a real-time PCR assay , as previously described [68] , but modified for measurement of mouse telomere [69] . PCR reactions were performed on the iCycler iQ5 2 . 0 Standard Edition Optical System ( Bio-Rad , Hercules , CA ) , using telomeric primers , primers for the reference control gene ( mouse 36B4 single copy gene ) and PCR settings as previously described [70] . For each PCR reaction , a standard curve was made by serial dilutions of known amounts of DNA . The telomere signal was normalized to the signal from the single copy gene to generate a T/S ratio indicative of relative telomere length . Equal amounts of DNA ( 20 ng ) were used for each reaction . The primers for telomere measurement by qPCR are listed in S2 Table . Telomere length and function ( telomere integrity and chromosome stability ) were estimated by telomere quantitative FISH [17 , 43] . Briefly , cells were incubated with 0 . 5 μg/ml nocodazole for 1 . 5 h to enrich cells at metaphases . Chromosome spreads were made by standard method . Metaphase-enriched cells were exposed to hypotonic treatment with 75 mM KCl solution , fixed with methanol: glacial acetic acid ( 3:1 ) and spread onto clean slides . Telomeres were denatured at 80°C for 3 min and hybridized with FITC-labeled telomere ( CCCTAA ) peptide nucleic acid ( PNA ) probe ( 0 . 5 μg/ml ) ( Panagene , Korea ) . Chromosomes were stained with 0 . 5 μg/ml DAPI . Fluorescence from chromosomes and telomeres was digitally imaged on a Zeiss microscope with fluorescein isothiocyanate ( FITC ) /DAPI filters , using AxioCam and AxioVision software 4 . 6 . Telomere length shown as telomere fluorescence intensity was integrated using the TFL-TELO program ( a gift kindly provided by Peter Lansdorp ) . Telomerase activity was measured by the Stretch PCR method according to the manufacturer’s instruction using TeloChaser Telomerase assay kit ( T0001 , MD Biotechnology ) . Briefly , about 2 . 5 × 104 cells from each sample were lysed . Lysis buffer served as negative controls . PCR products of cell lysates were separated on non-denaturing TBE-based 12% polyacrylamide gel electrophoresis and visualized by ethidium bromide staining . TRF analysis was performed using a commercial kit ( TeloTAGGG Telomere Length Assay , catalog no . 12209136001 , Roche Life Science ) . Cells were pretreated with RNaseA and Proteinase K ( PCR Grade , 03115879001 , Roche Life Science ) , followed by extraction using phenol: chloroform: isoamyl alcohol , digested with MboI ( R0147 , NEB ) at 37 °C overnight and electrophoresed through 1% agarose gels in 0 . 5 × TBE at 14 °C using a CHEF Mapper pulsed field electrophoresis system ( Bio-rad ) . Auto algorithm was used to separate DNA samples with a size range from 5 to 150 kb . The gel was blotted and probed using reagents in the kit . Tail or back skin tissues obtained from wild-type ( WT ) or G3 Terc deficient mice , or teratomas were fixed overnight in 3 . 7% paraformaldehyde at 4°C , dehydrated through graded alcohols and xylene , and embedded in paraffin . After deparaffinizing , rehydrating and washing in PBS , sections were incubated with 3% H2O2 for 10 min at room temperature to block endogenous peroxidase , subjected to high pressure antigen recovery sequentially in 0 . 01% citrate buffer for 3 min , blocked with 5% goat serum in PBS for 2 h at room temperature , and then incubated with the primary antibodies against K14 ( ab7800 , Abcam ) , Fst ( ab64490 , Abcam ) or P63 ( ab124762 , Abcam ) overnight at 4°C , washed and incubated for 2 h with appropriate fluorescence-conjugated secondary antibodies ( Goat anti Mouse IgG ( H+L ) , FITC , 115-095-003 , Jackson; Goat anti Rabbit IgG ( H+L ) , Alexa Fluor 594 , 111-585-003 , Jackson ) . For immunostaining of ES cells and in vitro differentiated cells , they were washed twice in PBS , then fixed in freshly prepared 3 . 7% paraformaldehyde in PBS ( pH 7 . 4 ) , permeabilized in 0 . 1% Triton X-100 ( Sigma–Aldrich , Saint Louis , MO ) in blocking solution ( 3% goat serum plus 0 . 5% BSA in PBS ) for 30 min , washed and left in blocking solution for 1 h . Cells were then incubated overnight at 4°C with primary antibodies and then secondary antibodies as described above . Blocking solution without the primary antibody served as negative control . Nuclei were counterstained with 0 . 5 μg/ml Hoechst 33342 in Vectashield mounting medium . Fluorescence was imaged using a Zeiss fluorescence microscope ( Axio Imager Z1 ) and using the same exposure time for each group . ImageJ software ( https://imagej . net/ ) was used for relative quantity measurement of fluorescence intensity . Region-of-interest ( ROI ) tool was used to select the cell or background , and the fluorescence intensity of ROIs achieved . Background with the same threshold was subtracted for each image . IF-FISH was performed based on an established protocol [71] . Briefly , immunostaining of the cells was performed as described above . After washing the excess of secondary antibody with PBS , cells were fixed in 4% formaldehyde for 2 min , dehydrated with ethanol , and incubated with FITC-telomeric PNA probe as described earlier for telomere QFISH . Fluorescence was imaged using the Zeiss fluorescence microscope . DNA methylation by bisulfite sequencing Genomic DNA was extracted from cells using DNeasy & Blood Tissue Kit ( Qiagen ) according to the manufacturer’s instructions . Bisulfite treatment of DNA was performed with the EpiTect Bisulfite Kit ( Qiagen ) . Bisulfite converted DNA was amplified by seminested PCR , using HS EX Taq DNA Polymerase ( Takara ) . Primer sequences are detailed in S3 Table . PCR products were recovered from stained gels ( EasyPure Quick Gel Extraction Kit , Transgen ) , cloned into a pEASY-T1 Simple Cloning vector ( Transgen ) and then sequenced . The plasmid pEASY-T1-Fst- overexpression ( OE ) and pEASY-T1-p63-OE were constructed by amplification of Fst or p63 cDNA by PCR and cloning it into pEASY-T1 simple cloning vector ( TransGen ) . Following digestion with XhoI and NotI , Fst or p63 sequences were inserted into Plch37 plasmid . Then the recombinant plasmids were transfected into J1 ES cells or MEF . At 48 h after transfection with 2 μg plasmid using lipofectamine 2000 ( Invitrogen ) , cells were collected for protein and RNA extraction . For obtaining stably transfected cell lines with Fst overexpression , cells were transfected with 2 μg plasmid using lipofectamine 2000 ( Invitrogen ) and selected by 1 . 5 μg/ml puromycin for 7–10 days , and clones were picked . shRNA sequences were synthesized ( S4 Table ) , and cloned into pSIREN-RetroQ , according to manufacturer’s instructions . The shRNAs without sequence homology to mouse genes served as a negative control . The RNAi retrovirus was packaged using Plat-E cells and then infected cells during differentiation . ChIP-qPCR analysis was performed as described previously [72] , with slight modification . Briefly , 5 × 107 cells were fixed with 1% paraformaldehyde , lysed , and sonicated to achieve the majority of DNA fragments with 100–1000 bp . DNA fragments were then enriched by immunoprecipitation with 5 μg H3K9me3 antibody ( ab8898 , Abcam ) , 7 μg H3K9Ac antibody ( ab4441 , Abcam ) , 5 μg Dnmt3b antibody ( ab13604 , Abcam ) , 5 μg H3K9me2 antibody ( ab1220 , Abcam ) or 5 μg H3K27me3 ( ab6002 , Abcam ) . The eluted protein:DNA complex was reverse-crosslinked at 65 °C overnight . DNA was recovered after proteinase and RNase A treatment . Real-time PCR was performed to compare the histone modification at the Fst promoter region using primers provided in S5 Table . Normal rabbit IgG ( #2729S , Cell Signaling ) or Mouse ( G3A1 ) mAb IgG1 Isotype Control ( 5415S , Cell Signaling ) served as negative control . Microarray was performed by CapitalBio Corporation ( Beijing , China ) using Affymetrix 430 2 . 0 oligonucleotide mouse arrays designed from GenBank , dbEST , and RefSeq sequences based on the UniGene database . The analysis was carried out based on the software and method provided by CapitalBio ( http://www . capitalbio . com ) . Only probe sets showing at least 1 . 8-fold change were retained in the final list . The detection call indicates whether a transcript was reliably detected ( P , Present ) or not ( A , Absent ) . We performed hierarchical clustering with the differentially expressed genes using cluster software ( version 3 ) and by applied mean centering and normalization of genes and arrays prior to average linkage clustering . The Fst promoter ( ~2000bp ) was cloned into pGL3-basic vector , following digestion with XhoI and HindIII . 2×105 ES cells per 12 well were transfected with 1 μg pGL3-basic vector containing Fst promoter and 10 ng pRL-SV40 vector as control using lipofectamine 2000 ( Invitrogen ) according to manufacturer’s instruction . 24 hours after transfection , ES cells were lysed with 1×PLB ( positive lysis buffer , Promega ) , shaken for 15 min , and then centrifuged at 13000 rpm for 10 min at 4°C . The supernatants were collected and analyzed for luciferase activity by dual reporter assay according to manufacturer’s instructions . XMP13 probe ( D-1413 , Metasystems ) specific for mouse Chromosome 13 was used for chromosome identification . FISH on chromosome spread was performed according to manufacturer’s instructions . The probe was added and coverslip placed , sealed with rubber cement , denatured by heating slide at 75°C for 2 min , and incubated in humidified chamber at 37°C overnight . Slides were washed and stained with 0 . 5 μg/ml DAPI in VectaShield antifade medium . Digital images were captured using a CCD camera on a Zeiss Imager Z2 microscope . The coordinates of the chromosome were recorded with the venire scale along the top and side of the microscope stage . The slides were washed and performed with telomere FISH as described above . After staining with DAPI again , fluorescence from chromosomes and telomeres was digitally imaged using the same microscope according to the recorded coordinates . The telomeres of chromosome 13 were revealed by comparison of the images from the same coordinates . The data from multiple groups were analyzed by ANOVA , and means were compared by Fisher’s protected least significant difference ( PLSD ) using the StatView software from SAS Institute . T-test was used to analyze statistical significance of the two-paired groups . Significant differences were defined as p < 0 . 05 , 0 . 01 , or lower . | Patients with mutations in the telomerase component ( eg , Dyskerin , TERT , TERC ) are frequently accompanied by symptoms of abnormal epidermis , such as hyperpigmentation , premature skin degradation , hair follicle shedding , skin atrophy , and dry skin . Mice with mutations in telomere-associated proteins or telomerase genes also show similar phenotypes , associated with telomere shortening . However , the underlying molecular signaling and mechanisms remain elusive . Here , we show that the differentiation of epidermis is disrupted resulting from short telomeres . Epidermal differentiation abnormalities can be rescued as the telomere length is extended . Furthermore , we uncover that Fst-BMP-Smad pathway is implicated in regulation of epidermal differentiation by telomeres length . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"skin",
"medicine",
"and",
"health",
"sciences",
"integumentary",
"system",
"chromosome",
"structure",
"and",
"function",
"cell",
"differentiation",
"telomeres",
"developmental",
"biology",
"stem",
"cells",
"epidermis",
"epigenetics",
"dna",
"chromatin",
"dna",
"methyla... | 2019 | Telomere dysfunction impairs epidermal stem cell specification and differentiation by disrupting BMP/pSmad/P63 signaling |
Systematic , genome-wide RNA interference ( RNAi ) analysis is a powerful approach to identify gene functions that support or modulate selected biological processes . An emerging challenge shared with some other genome-wide approaches is that independent RNAi studies often show limited agreement in their lists of implicated genes . To better understand this , we analyzed four genome-wide RNAi studies that identified host genes involved in influenza virus replication . These studies collectively identified and validated the roles of 614 cell genes , but pair-wise overlap among the four gene lists was only 3% to 15% ( average 6 . 7% ) . However , a number of functional categories were overrepresented in multiple studies . The pair-wise overlap of these enriched-category lists was high , ∼19% , implying more agreement among studies than apparent at the gene level . Probing this further , we found that the gene lists implicated by independent studies were highly connected in interacting networks by independent functional measures such as protein-protein interactions , at rates significantly higher than predicted by chance . We also developed a general , model-based approach to gauge the effects of false-positive and false-negative factors and to estimate , from a limited number of studies , the total number of genes involved in a process . For influenza virus replication , this novel statistical approach estimates the total number of cell genes involved to be ∼2 , 800 . This and multiple other aspects of our experimental and computational results imply that , when following good quality control practices , the low overlap between studies is primarily due to false negatives rather than false-positive gene identifications . These results and methods have implications for and applications to multiple forms of genome-wide analysis .
RNA interference ( RNAi ) is a gene-specific silencing process directed by short double stranded RNAs or small interfering RNAs ( siRNAs ) that can “knock down” expression of a selected gene by inducing messenger RNA ( mRNA ) degradation in a sequence-specific manner [1] . RNAi has been widely used as a molecular tool to selectively inhibit the expression of a chosen gene . By expanding this technique to use large-scale RNAi libraries , high-throughput RNAi analysis has become a powerful approach to screen essentially all genes of an organism , to identify gene functions that support or modulate any biological process of interest . Genome-wide RNAi analyses have been used to study many important biological processes and have provided novel , key insights . One important application of genome-wide RNAi screening has been to identify host genes that are required for the replication of a particular virus [2]–[4] . In several cases , two or more independent RNAi screens have been performed to identify host factors required by the same viruses [5]–[11] . An emerging challenge shared with some other genome-wide approaches is that such independent genome-wide RNAi studies often exhibit limited overlap in the lists of genes implicated . For example , there is only 3–6% overlap of gene lists identified from three genome wide RNAi studies for host factors of HIV [6] , [7] , [12] , [13] . Lack of overlap between studies must be due to some combination of false-positive and false-negative factors . If the dissimilarity between the sets of identified host factors is predominantly from false-positive factors , the majority of the genes identified from a genome wide RNAi study would be false positives . In this case , the number of genes involved in a biological process such as virus infection would be low relative to the number implicated . On the other hand , a high false-negative rate could also induce low agreement , since in this case each study would confirm just a small subset of involved genes , and many more genes than one study could identify might be implicated in the process . A critical step to reduce false discoveries is good quality control , which is essential due to the limited number of repeats possible for genome-wide analyses , frequently high assay noise , the potential for systematic errors , and other effects . Relevant quality control practices include adequately designed positive and negative controls , proper patterning of samples and controls on microtiter assay plates , data display approaches suitable for revealing systematic errors , etc . Nevertheless , even good quality control cannot eliminate error , so that quality control is an essential but not sufficient basis for effective RNAi analysis . Appropriate data analysis , such as false discovery rate ( FDR ) controlling procedures , should also help to minimize false discoveries [14] . However , the FDR is controlled within each study , and is not designed to control false-positive error rates caused by sources of variation between studies . Moreover , an inherent problem for large-scale dataset processing is that a strict cut-off to limit false-positive results increases the number of false negative results , and vice versa [1] . Depending on the goal of a study , the tolerance for false positives or false negatives might be adjusted accordingly to generate the final results . Of course , data analysis is not the sole source or solution for false-positive and-negative results . Both technical and biological sources also contribute to false discoveries and non-discoveries in genome-wide RNAi studies . An siRNA may silence one or more genes besides the targeted one [15] , owing to incomplete sensitivity and specificity of siRNAs [16]–[19] . Beyond these off-target effects , another false-positive factor is measurement error intrinsic to the complex phenotypic readouts typically used . Similarly , many experimental issues contribute to false negatives . For example , the organism under study may have genetic redundancies that limit the accessibility of certain functions to phenotypic manipulation by knocking down a single gene [20] , [21] . Furthermore , genes with undetectable expression or whose knock down results in cytotoxicity are usually excluded from such RNAi analyses , further reducing the number of genes that are accessed in such “genome-wide” studies [22] . Besides the above issues , false negatives also can be generated by systematic errors such as plate position effects [23] , by failure to control for variation in local cell environments [24] , [25] , by inefficiencies in knocking down targeted genes [26] , and by other effects . For these and other reasons noted above , use of well-chosen quality control as well as analytic methods are critical for producing high quality screening results [23] , [27] . To better understand RNAi screening and the relative contributions of false positives and false negatives , we performed a meta-analysis of four recent studies to identify host genes involved in the replication of influenza virus , an important human pathogen [8]–[11] , [28] . Despite differences in the RNAi libraries and cell lines used , these studies , including one from our laboratory [9] , employed similar two-step approaches . All studies began with a high-throughput primary screen with an RNAi library targeting the whole genome . Candidate genes from this primary screen then were re-tested for function in virus replication in repeated secondary validation assays with individual siRNAs . Similar to prior HIV results , the pair-wise overlaps among the confirmed gene lists were only 3–15% ( Figure 1 ) . Using a multi-faceted meta-analysis , we found that the independent gene lists are strongly interlinked by functional pathways and protein-protein interactions , and that their low overlap is due primarily to false negative rather than false-positive factors . First , there is substantially higher agreement between studies from the perspective of functional categories rather than gene lists . Second , the combined list of genes confirmed by all studies embodies a much richer network of molecular interactions than expected by chance . Thus , genes identified in independent RNAi studies are physically as well as functionally connected . Finally , we developed a new statistical model that incorporates the main experimental features of RNAi screening and that , upon fitting to the data via likelihood and Bayesian techniques , estimates the major intrinsic parameters governing false positives and false negatives . This model well duplicates the statistical patterns in the multi-study gene level data , and indicates that low overlap arises primarily from false negative factors . Thus , our bioinformatic and statistical analyses show that current genome-wide RNAi screens each reveal a highly useful but partial glimpse of a larger whole .
Gene-level results from primary and secondary screens were collected from four genome-wide RNAi screens to identify host factors crucial for influenza virus replication [8]–[11] . All human genes were mapped to Entrez IDs ( NCBI , http://www . ncbi . nlm . nih . gov/ ) ; human orthologs for Drosophila genes were extracted from www . ensembl . org . Table S1 lists for each study the Entrez IDs of all confirmed genes , followed by unconfirmed primary candidates . Study-specific lists of confirmed genes were compared using a mean overlap fraction ( MOF ) . On average over all pairings of lists , MOF is the mean value of the proportion of one list that overlaps with another . Gene-set analysis was performed using confirmed genes from each of the four studies and functional-category information from the Gene Ontology ( GO ) project , as accessed using the R and Bioconductor systems ( www . bioconductor . org ) . We applied the recently-developed multi-function analyzer ( MFA ) [29] , [30] to process each gene list ( four study-specific gene lists and one combined list ) . Briefly , MFA provides a model-based analysis of a gene list . A gene's presence on the list is explained by latent binary activities of GO terms which annotate it; through model-based computations MFA determines the posterior activity probabilities over all terms in the collection . It refines model-based gene set analysis ( MGSA ) [31] by encoding natural constraints on the function-level activities in order to improve statistical efficiency . Like MGSA , MFA addresses deficiencies of simpler gene-set methods ( e . g . , in handling set overlaps ) by the simultaneous analysis of all sets in a collection . Gene-permutation was used to calibrate the between-study agreement of MFA-derived set lists . Specifically , data were organized in an incidence matrix with rows for genes and columns for studies , and containing indicators of confirmation . Permutation proceeded by shuffling row labels ( gene IDs ) and retaining the numbers of genes co-confirmed in two , three , and four studies , and expressed the null assumption of no functional association between studies . We computed Monte Carlo p-values using the MFA results from the observed data and from 999 permuted versions . Computations used GO . db version 2 . 8 . 0; and org . Hs . eg . db version 2 . 8 . 0 ( September , 2012 ) . In total 2472 GO terms were used; these are all terms that have non-empty intersection with the combined list of influenza-related genes and that are moderate in size ( between 5 and 50 human genes ) . The size constraint improved computational efficiency without much loss in functional information . Analysis of molecular interactions among the confirmed genes was performed in order to further assess the relatedness of the genes , both within studies and across studies . We assembled interaction graphs; these are data structures in which the vertices correspond to genes and the edges correspond to protein-protein interactions . The set of interactions used consisted of 47 , 647 human , protein-protein interactions from the BioGRID database [9] . A connected component is a maximal subgraph in which any two genes in the subgraph are connected by one or more paths . To assess the extent of inter-relatedness within a list of confirmed genes , we considered several properties of the subgraph that results from selecting only the confirmed genes and their impinging interactions . First , we measured the number of edges in the resulting connected components and the size ( in terms of genes ) of the largest connected component . Second , we measured the average degree of the vertices in the confirmed-gene subgraph . The degree of a vertex in a graph is the number of other vertices to which it has edges . Thus we measured , on average , the number of other confirmed genes with which each confirmed gene has known interactions . To determine if any of these measures is surprisingly large , we employed a Monte Carlo test in which the null hypothesis is that the measure can be accounted for by a randomly selected set of pseudo “confirmed” lists . Given a set of n actual confirmed genes , this test involved repeated random selection of gene sets of size n . We selected these random gene sets such that the degree of connectivity of the selected genes , with respect to the entire interaction graph , was the same as the degree of connectivity of the given list of confirmed genes . That is , if there are three genes in the confirmed list that each has eight known interactions with other genes in the genome , then our randomly selected gene sets would also include exactly three genes with eight interactions each . Note that the degree of connectivity we consider when doing this selection process refers to the connectivity of a given gene to all genes in the genome , as opposed to its connectivity to other confirmed genes . For the results reported here , our Monte Carlo p-values involved 9 , 999 iterations . To determine if there were more relationships between pairs of genes confirmed in different studies than would be expected by chance , we pooled the confirmed genes from the four independent studies and determined the connected components that resulted from this set of pooled genes . As before , we counted the number of edges in connected components , determined the size of the largest connected component , determined the average degree for confirmed genes , and assessed the statistical significance of each measure using a Monte Carlo methodology . Additionally , to assess the extent to which the genes confirmed in separate studies were related , we counted the number of spanning edges in the connected components . A spanning edge is one that represents an interaction between two genes that were confirmed in different studies . We used a Monte Carlo test to measure the statistical significance of the number of spanning edges we observed . Each iteration of the Monte Carlo test involved randomly selecting , for each study in the pool , a set of pseudo “confirmed” genes which have the same degree of connectivity as the actual confirmed genes in that study . Given these randomly chosen sets , we counted the number of spanning edges in each as we did with the actual data . Again , our Monte Carlo p-values used 9 , 999 iterations . To provide reasoned inferences about factors affecting among-study gene-level agreement , we developed a statistical model for genome-wide RNAi studies and corresponding likelihood-based analysis methods . The model formulates relationships among: system-level parameters that affect sensitivity and various error rates , gene-level and study-level latent variables that transduce information about the system to information at the gene-level , and gene-level , multi-study data on both detection and confirmation by RNAi screening . In its generative form , the model specifies the probability of observing any particular multi-study data set . In its inferential form , it indicates the likelihood assigned to any particular parameter setting in light of observed data [32] .
Four nearly genome wide RNAi studies have been published to identify host genes that affect influenza virus replication ( summarized in Table 1 and 2 ) . These four studies differed in the cell lines and RNAi libraries used , but each invoked a two-stage screening strategy in which candidate genes detected in a primary , genome-wide screen ( Table 1 ) were subjected to more thorough testing in a second , validation phase ( Table 2 ) . In such validation testing , the candidate genes were knocked down using an alternate dsRNA or multiple single siRNAs to confirm that the effects on influenza virus were not due to off-target effects on unintended genes . The majority of genes detected and confirmed in these four genome-wide screens are genes that promote influenza virus replication , i . e . , knock-down of these genes by RNAi decreased virus replication ( Table 2 ) . Two screens using human A549 cells identified , respectively , 219 and 168 genes that promoted influenza virus replication , without reporting any host genes that restrict influenza virus - i . e . , genes whose knockdown increased viral replication . One screen in human U2OS cells found 129 genes that promoted influenza virus replication , and 4 genes that restricted replication . A screen in Drosophila DL1 cells identified 104 genes that promoted virus replication and 11 genes that restricted virus replication . Of the 104 Drosophila genes that promoted influenza virus replication , 96 have a total of 154 human homologs according to the Ensembl database , while 10 out of the 11 Drosophila genes that restrict influenza virus have 14 human homologs . As these studies identified very few genes that restricted virus replication when knocked down , our analysis focused on the genes that promoted virus replication . From 984 unique human genes identified in the primary screens of these four studies , 614 unique genes were confirmed that promoted influenza virus replication . The symbols and Entrez ID numbers of all 984 genes are listed in Supplemental Table S1 , with unconfirmed genes at the bottom of the table in shaded rows . Thus , on average , each such study detects approximately 1% of the genes in the genome as potentially involved in the influenza virus replication , and confirms approximately half of these candidate genes . Although all four studies aimed to perform a general identification of host genes affecting influenza virus replication , their gene lists exhibited relatively little overlap . Only one gene ( COPG , or coatomer protein complex , subunit gamma ) was detected and confirmed by all four studies; nine genes were confirmed by three of four studies , and 35 by two studies ( Figure 1A ) . Pairwise overlap between studies ranged from 3% to 15% , with a mean pairwise overlap of 6 . 7% ( Figure 1B ) . To complement comparisons at the gene level , we examined the relationship of the four confirmed-gene lists at the level of functional categories recorded by the Gene Ontology ( GO ) project . Within-study lists of over-represented GO terms exhibited relatively strong agreement ( Figure 4 ) , with a mean overlap fraction of 19% ( compared to 6 . 7% for gene lists ) . This agreement is substantially more than would be expected in the absence of functional associations between the studies ( Monte Carlo p-value = 0 . 001 ) . The finding is based on an advanced gene-set analysis tool which accommodates term-size and term-overlap issues ( MFA; see Methods ) ; the same conclusion was found with simpler gene-set enrichment methods of GO-term lists suggests that the four studies were probing common functional signals in influenza virus dependence on host genes . The functional categories over-represented in the combined list of 614 confirmed genes from all four studies represent a wide variety of functions . The top 29 categories are illustrated in Figure 5 , including with categories associated with mRNA translation ( ribosomal small subunit , eIF3 initiation complex , ribosome binding ) , vesicular transport ( Golgi to ER , vacuolar ATPse ) RNA metabolism ( RNA splicing , transport , poly ( A ) regulation ) , regulated protein degradation ( proteasome , ubiquitination factors ) , and other functions . Cellular functions like regulation of type I interferon production and nucleo-cytoplasmic transport are known to play important roles in influenza virus replication , while other functions like NADP binding and vitamin transporter activities are novel findings identified through our analysis . Figure 5 also shows that , for many categories , two or more studies isolated distinct sets of genes in the same functional category , so that the same cellular functions were independently but repeatedly identified to promote influenza virus replication . As noted above , if many of the genes confirmed in the genome-wide influenza screens were false positives , then we would expect these genes to be distributed across unrelated pathways and functional complexes . That is , we would expect relatively few direct interactions among the genes confirmed in one study or in different studies . Given the confirmed genes from the four influenza screens , we used known physical interactions among the genes' products , as cataloged in the BioGRID protein-protein interaction database [9] , to test two specific hypotheses: First , that the genes confirmed within each individual study are more inter-related by protein interactions than would be expected by chance . Second , that there are more such relationships between genes confirmed in different studies than would be expected by chance . To assess the relatedness of a set of confirmed genes , we first determined the number of connected components in the interaction subgraph consisting of the proteins encoded by the confirmed genes . We considered two properties of these connected components: the number of edges in the connected components and the number of genes in the largest connected component . We further measure relatedness by determining the average degree for vertices in the subgraph of confirmed genes . We also investigated augmenting our interaction data with metabolic-pathway and transcription-factor relationships , and found that the results of our analysis were qualitatively similar when we include these additional interaction sets . To test our first hypothesis , we independently measured the above mentioned properties for the four interaction subgraphs constructed using the confirmed genes from each individual study . To determine if any of the measures was surprisingly large , we employed a Monte Carlo test with the null hypothesis that these measures could be accounted for by a randomly selected set of pseudo “confirmed” genes . Given a list of n actual confirmed genes , our Monte Carlo test involved repeated random selection of gene sets of size n . We selected these random gene sets such that the degree of connectivity of the selected genes , with respect to the entire interaction graph , was the same as the degree of connectivity of the given list of confirmed genes . The rationale for controlling the degree of connectivity in our Monte Carlo tests is that we want to rule out the possibility that a confirmed list has a high degree of connectivity with other confirmed genes simply because its genes have many known interactions . Figure 6A illustrates the two connected-component measures and the average degree measure , and shows the counts and p-values calculated using the graphs for each individual study . For three of the four studies , the number of interactions in connected components and the average degree exceeded those of the null hypothesis by a statistically significant margin ( p<0 . 05 ) , and for two of the studies , the number of confirmed genes in the largest connected component is statistically significant . These results generally support our first hypothesis that the confirmed genes have a higher degree of relatedness than expected by chance . To test our second hypothesis , we pooled the confirmed genes from the four independent studies and determined the connected components and the average degree in the interaction subgraph for this set of pooled genes . Figure 6A ( bottom row ) shows the resulting number of edges in the connected components , the size of the largest connected component , and the average degree for vertices in the pooled subgraph . The number of protein interactions in the pooled study subgraph ( 214 ) is much larger than the sum of interactions in the individual-study graphs ( 121 ) , and was highly statistically significant at p<0 . 0001 . The average degree of proteins in the pooled subgraph is larger than the average degree of any of the single-study subgraphs , and this degree was highly statistically significant ( p<0 . 0001 ) . Thus , these results strongly support our second hypothesis that there are more interactions between pairs of genes confirmed in different studies than expected by chance . Likewise , the size of the largest connected component in the pooled graph ( 108 genes ) is much larger than the sum of the largest connected components in the individual-study graphs ( 34 genes ) and was statistically significant at p<0 . 05 . Figure 7 illustrates the 108-gene largest connected component from the pooled-study graph . As this figure indicates , the genes confirmed by each individual study tend not to be topologically clustered , but instead generally have interactions with genes confirmed by other studies . To further probe our second hypothesis , we counted the number of spanning edges in the connected components of the pooled-study graph . A spanning edge is one that represents an interaction between two genes that were confirmed in different studies ( Figure 6B ) . As before , we measured the statistical significance of the observed number of spanning edges by iterative Monte Carlo tests that involved randomly selecting , for each study in the pool , a set of pseudo “confirmed” genes with the same degree of connectivity as the actual confirmed genes in that study . The results show that , of the 214 total protein-protein interactions of the pooled study subgraph ( Figure 6A ) , 175 ( 82% ) were spanning interactions between gene products confirmed in different studies ( Figure 6B ) . This result was statistically significant at p<0 . 05 . We also considered the possibility that our measures of interactivity might be inflated by the fact that some of the genes confirmed in the DL1 study mapped to multiple human orthologs . To control for this effect , we repeated all of our analyses using a smaller set of DL1 genes in which each of the 96 confirmed Drosophila genes was mapped to exactly one human gene . In particular , we mapped each Drosophila gene to the human ortholog with the fewest known interactors . The differences in p-values between the original analysis and this “conservative ortholog mapping” re-analysis were all small . Specifically , none of the p-values changed from statistically significant ( p<0 . 05 ) to not significant . Overall , then , the results show that the independent RNAi studies identified distinct but physically interacting sets of genes , and that these confirmed gene products exhibit significantly more interactions both within and between studies than expected by chance . Multiple independent analyses above show that the outputs of the four separate screens are not randomly divergent , but are highly interlinked both physically and functionally . To more thoroughly examine the biological and experimental factors that affect agreement among such screens , we generated a statistical model of RNAi screening that incorporates known biological and experimental features to better analyze and understand the expected relationships between the outputs of independent screens . The assessment of agreements and disagreements among studies has long been a focus of model-based statistical analysis , from seminal work by R . A . Fisher and colleagues on species abundance estimation in ecology [34] , through more recent and relevant precursors [35]–[37] to our own calculations . The rationale for this general approach is that the specific findings of any study are affected by numerous factors , some of which are systematic and shared in some predictable way among studies , and some of which are idiosyncratic . To capture the systematic effects we treat them as parameters in a stochastic process presumed to have generated the observed data , and we infer the parameter values by calculating the probability of observed data ( the likelihood ) . The structure of RNAi experiments forces us to go beyond previously described probability models and propose a specification for multi-study , two-stage ( detection/confirmation ) , genome-wide RNAi data . As described more fully in the Methods and Text S1 , the model considers that a fraction in ( 0 , 1 ) of all G genes is involved in the phenotype in the sense that error-free measurements in some cell type would detect and confirm this involvement . However , both primary detections and secondary confirmations are subject to additional sources of variation . On average over the full set of siRNAs , there is a knockdown threshold parameter in the range ( 0 , 1 ) indicating the frequency with which a single siRNA would reduce expression and function of a targeted gene sufficiently to cause detection or confirmation if the gene were involved and if knockdown and phenotypic readout were free of errors and other interfering effects ( see next paragraph ) . Since siRNA pools show greater phenotypic penetration than an individual siRNA [38] , the model further specifies how detection/confirmation rates increase as we apply more siRNAs targeting the same gene ( see Methods ) . Figure 2A illustrates the conditional independence relationships of observable , reported data and latent variables in this statistical model , while Figure 2B shows the nature of the distributions that were assumed for each variable . As a separate issue from the frequency of phenotypically significant knockdown at the mRNA level , we allow that many factors could block a study from detecting a phenotype associated with an involved gene . We lump these factors into study-specific accessibility parameters , which accommodate many factors specific to cell types and conditions used in different studies . Examples of such potentially confounding factors include whether the range of genes targeted by the siRNA library used; whether the cells and conditions used for a targeted mRNA is not expressed; if the mRNA is too abundant to be cleared within the period of RNAi treatment; if the functional protein product is too long-lived to be depleted over the measurement time course; if knockdown induces general cytotoxicity , which is usually taken to disqualify scoring for specific involvement; or if the phenotypic effect of knocking down an involved gene is masked by functional redundancies such as expressed homologs , parallel pathways , etc . . All other false negative measurement errors are monitored by type II error parameters , which were also made study specific since this improved model fit . Finally , the model allows false-positive detections or confirmations to occur either because of off-target effects ( the parameter is the mean number of off-targets per siRNA ) or by type I measurement error ( parameter ) . Numerical and Monte Carlo methods were used to obtain parameter estimates and approximate confidence intervals using only a set of modeling assumptions and the multi-study detection/confirmation records ( Methods and Supplementary Text S1 ) . We computed three simulation diagnostics to check the internal validity of the model-based inferences , as detailed in Supplementary Text S2 , Section 2 . First we performed a consistency check to assure the integrity of the modeling calculations and their implementation in software ( Text S2-2 . 1 ) . Likelihood theory predicts that maximum likelihood parameter estimates computed from ever increasing data sets converge to the underlying parameter settings when the model is identifiable . Thus , we simulated gene-level data from the model , under various parameter settings , and demonstrated that as the genome size increases these settings were recovered . Second , we performed predictive checks to examine properties of synthetic data generated from the fitted probability model ( Text S2-2 . 2 ) . Table S2 in Text S2 demonstrated that simulations of the estimated generative model recapitulate statistical patterns seen in the observed data . For example , Figure S2 in Text S2 plots the number of confirmations against the number of detections , both in the observed data and in hypothetical repeats . The goodness-of-fit is compelling in this plot and in several other summaries ( Table S2 in Text S2; Figure S3 in Text S2 ) . For the third check , we did leave-one-study-out diagnostics . We fitted the model four times , once to each subset of three studies obtained by leaving out a single study's results ( Supplementary Text S2-2 . 3 ) . The cross-validation findings indicate stability of the parameter estimates ( Table S3 in Text S2 ) as well as accuracy in predictions of the left-out studies ( Table S4 in Text S2 ) . We plotted and compared the distribution of gene counts across the 81 possible patterns ( Table 3 ) between simulated and experimental data , and the simulation results generated from model fits accurately reflected the experimental data ( Figure S7 in Text S2 ) . Therefore , our model faithfully simulates actual RNAi studies . Table 4 reports all parameter estimates in the fitted model . The threshold parameter for the frequency of effective knockdown at the mRNA level was estimated to be 75–99% ( 95% CI ) , which indicates that there is a fairly high chance for an involved gene to be scored as positive , if the gene is accessible and measured in an error free system . The largest factor affecting inter-study agreement was the general accessibility rate , which we estimated separately for each study to improve model fit , and which ranged from 5–17% ( 95% CI across all studies ) . In other words , these results imply that , in a typical genome-wide RNAi study , most influenza virus-involved genes are inaccessible because , as noted above , one or more of many potential confounding factors either blocks knockdown or interrupts the transfer of this effect into a measurable phenotype . Biological factors and measurements consistent with such limited genetic accessibility are considered in the Discussion . Another study-specific error causing false negatives is type II measurement error , which depending on individual study setup contributes 1–50% false negatives ( 95% CI across all studies ) . By contrast , both parameters dictating the frequency of false positives were estimated to be low . The type I measurement error is estimated at 0 . 3% , which corresponds to approximately one false-positive well per 384-well plate . The off-target rate is further discussed in the next section . Besides type I measurement error , false positives can also result from knocking down an unintended gene during RNAi , i . e . , the off-target effect . In our model estimation , the off-target rate was lower than initially expected , with an estimated average of up to 0 . 032 off-target genes per siRNA . We wondered if this finding might be attributable to our use of a Poisson distribution model for the number of off-targets per siRNA and performed robustness checks to test this assumption ( Supplementary Text S2-2 . 4 ) . Little is known about the details of off-target distributions , although computational predictions from a Drosophila dsRNA library suggest a highly non-Poisson structure [33] . In fact , the Poisson assumption made little difference for parameter estimation , as evidenced by computations done under a highly over-dispersed negative-binomial alternative to the Poisson ( Supplementary Text S2-2 . 4 ) . The multi-study data prefer a small off-target rate , as shown by the profile likelihood plot for this parameter in Figure 8A . In this calculation , we fixed at various sub-optimal values and computed maximum likelihood estimates of the remaining parameters using the multi-study RNAi data . A source for the lack of fit is shown in Figure 8B: increasing above its maximum likelihood estimate the model cannot explain the relatively high observed confirmation rate among detected candidate genes ( mean rate over 50% ) . Figure 8C shows the effect on key error rates in fixing at sub-optimal values . False-positive and false-negative rates can be defined in various ways , depending on the reference set of genes . Using estimates and uncertainties in the model parameters , we estimated the false discovery rate ( FDR ) , the false non-discovery rate ( FNDR ) , the false positive rate ( FP ) and the false negative rate ( FN ) , and we obtained posterior distributions for each in order to get approximate confidence intervals ( Figure 9 ) . FDR , which is widely used in high-throughput studies , is the rate of false positives among all statistically significant findings . In the case of the influenza study , the FDR is the probability that a gene is not involved with influenza given that this gene was confirmed in a secondary screen to be involved ( i . e . , the denominator counts the number of confirmations ) . Similarly FNDR is the probability of involvement given the gene is not confirmed ( whether or not it is detected ) . For design purposes it is often useful to think of errors relative to the true set of involved genes ( FN ) or non-involved genes ( FP ) . By either sets of measures , the clear indication from Figure 9 is that false-negative factors dominate false-positive factors in explaining the limited agreement among studies . Thus , the low overlap among the confirmed gene lists of the different studies arises principally due to missed genes . Since the above results infer that the low gene level overlap between and among the four documented studies ( Figure 1 ) is primarily due to false negatives ( Figure 9 ) , the total number of involved genes is expected to be significantly more than the 614 confirmed in these initial studies . Our model estimated the rate of gene involvement in influenza virus replication ( ) at 12% of the genome . Taking G = 22 , 000 total genes , this corresponds to N = 2766 genes , with a 95% posterior confidence interval of ( 2306 , 3342 ) . Supplementary Text S2-4 and Figure S8 in Text S2 provide further details on this inference . Figure 10 shows the predicted progression in additional confirmations beyond the 614 initial genes if further RNAi studies of similar in design were performed ( details in Supplementary Text S2-4 ) . As illustrated in Figure 10 , a total of ∼12 studies would be required to identify the majority of host genes involved in influenza virus replication .
Although individual genome-wide RNAi studies typically implicate from one to several hundred genes in a given biological process , multiple independent findings in our results indicate that such studies typically miss most involved genes due to a high false-negative rate during primary genome-wide screening . Thus , the results of each study provide a partial glimpse into a larger , interconnected whole . For the influenza virus studies examined here , e . g . , although the gene level agreement between RNAi studies is relatively low ( mean overlap fraction 6 . 7% ) , much higher levels of agreement between studies ( 19% ) exist at the level of functional categories ( Figure 4 ) . Further , we found that the gene sets implicated in independent studies are highly connected by interactions among their protein products ( Figures 6–7 ) . These analyses examined the implicated genes from different perspectives , but all indicated that the gene sets identified in independent studies are significantly related . We conclude that , in general , functional pathways are better represented than individual genes in genome-wide RNAi studies , that independent RNAi studies add value , and that single RNAi studies can be valuable , particularly if interpreted in light of these and other insights . Using the measurements from the four influenza virus studies , we developed a statistical model to estimate critical parameters that control the output of genome-wide RNAi studies . The model was designed to reflect the experimental process of RNAi screening and to utilize the available published information from such screens . Our model was refined through several rounds of iteration and fits the experimental data very well ( Table S2 in Text S2 , Figure S2 , S3 and S7 in Text S2 ) . The leave-one-out test showed that the range and medians of the model estimations match the real experimental data well ( Table S3 in Text S2 ) . In addition to other insights , our statistical modeling confirmed that an individual genome-wide RNAi study generally will only identify a small portion of genes implicated in a biological process . The model further indicates that this partial coverage results predominantly from false-negative errors in the high-throughput , genome-wide primary gene detection phase of the analysis , and that false positives in the more focused , repeated validation phase of analysis are much rarer . Below we discuss factors related to these false-positive and false-negative rates . As noted above , many issues might lead to false discoveries in a genome-wide high-throughput screen . However , to reduce false discoveries , all four studies in our meta-analysis incorporated validation testing of implicated genes with independent siRNAs and multiple repeats . Stringent statistics were used to control the false discovery rate , normally to below 5% . Accordingly , such validated genes , although only confirmed in one study , are unlikely to be false positives . A potentially significant issue causing false positives for any RNAi knockdown study ( either single gene or genome-wide ) is the possibility of off-target effects [15] , [16] , [33] , [39] , [40] . In most genome-wide RNAi studies , including the four analyzed here , the potential for off-target effects in the initial screening phase is addressed by requiring the knock-down phenotype to be independently confirmed by independent , usually multiple tests with two or more distinct siRNAs against the target gene . As siRNAs with distinct sequences are unlikely to affect overlapping off-target genes , such testing considerably reduces the potential for off-target induced false positives . Indeed , the estimated off target rate was low , according to model based calculations ( Table 4 ) . Our further analysis showed that if we forced the off-target rate to be higher , the model fit the data poorly and did not explain the observed confirmation rates of >50% of the detected genes . Increasing the off target rate to ∼5 genes/siRNA , e . g . , shifts the estimated confirmation rate to 5% or less , which contradicts the data from all four studies ( Figure 8B ) , and amplifies the estimated gene involvement rate to over 80% , which appears unlikely from many practical and biological considerations . The actual model-estimated rate of host gene involvement of 12% ( Table 4 , ) , while higher than predicted from individual studies , is consistent with the many empirical and a priori considerations supporting the conclusion that true negatives should significantly outnumber true positives in such screens for viral dependency factors . By comparison , the sheer scale and technical challenges of genome-wide studies presently make false negatives hard to control in the initial detection phase of global RNAi studies . Because of the large number of samples to test ( typically tens of thousands ) , only limited replicates ( typically 2 or 3 ) can be performed in the primary , genome-wide screens . Furthermore , simultaneous processing of hundreds to thousands of samples in microwell plates makes it impossible to individually optimize many critical parameters , such as assay timing , for knockdown analysis of each gene . Moreover , choices made to facilitate such high-throughput analysis , such as tractable endpoints , assays or virus genotypes may further affect the ability to detect at least some phenotypic effects . These and other factors noted below , including genetic redundancies , can restrict detection sensitivities and reduce phenotypic impacts , jeopardizing the detection of potentially positive phenotypes against the background of often significant experimental noise in such large-scale , high-throughput measurements . Indeed , relative to most cell-based screens , screens for effects on viral replication are subject to additional variability associated with fluctuations in the efficiencies of initial and later stages of infection . Notably , substantial portions of such variations in infection efficiency are correlated with cell density , cell size and other local population features of the cultured cells , and appropriate normalization for such factors can significantly improve the consistency of the assay results [24] , [25] . In addition to such within-screen variability , and as noted in part above , variability between independent screens is increased by differences in viral genotypes , cells , culture conditions , assay design ( partial or full virus lifecycle , direct or indirect readout , timing , etc . ) , and similar experimental issues ( Table 1 ) . This combination of issues , plus additional sources of variable and systematic errors noted in the Introduction , makes the use of appropriate quality control practices and analytical methods [27] particularly critical for virus-oriented screens . Besides the above technical challenges , our statistical inference indicates that a major contributing issue to false negatives is low gene accessibility , i . e . , the portion of genes for which the method of a given RNAi study can show a corresponding loss-of-function phenotype if the gene is involved . Specifically , our fitted statistical model estimated from the data of the four influenza virus studies that only ∼10% of tested genes could be effectively assayed by RNAi knockdown for phenotypes in any one study ( Table 4 ) . This statistically inferred low gene accessibility is consistent with and presumably results from multiple aspects of host cell biology . First , genes not expressed in the specific cell type used for an RNAi study will not show any phenotype upon attempted knocked down . In analyses of gene expression in 84 human tissues and cell lines using high coverage Affymetrix DNA microarrays , only 37% of total probes were found to have moderate or higher expression in at least one cell type [41] . These probes mapped to 9214 different genes , and the number of genes expressed in each cell type ranged from 484 ( ovary tissue ) to 6038 ( B lymphoblast ) . In lung tissue , relevant to influenza virus infection , only 3355 genes , or ∼15% of the total were expressed [41] . Thus , in many cell types , a relatively small fraction of genes may exhibit sufficiently high expression to produce measurable phenotypic effects . Moreover , the particular sets of genes that are functionally expressed vary between different tissue and cell types in vivo , and between different cells and cell lines in culture [41] . Accordingly , RNAi screening in different cell lines will reveal contributions from different genes , either directly because of differences in the expression of particular tested genes or indirectly , because the contribution of a tested gene can be masked by other complementing genes that are differentially expressed between the cell lines ( see also below ) . Second , in genome-wide RNAi studies , genes that cause cytoxicity when knocked down are generally excluded from analysis . In the A549DE study [8] , e . g . , 1520 siRNAs were judged to be cytotoxic and were omitted from further analysis . Third , as cells have been selected for functional robustness , knocking down expression of a single gene may have little phenotypic effect when additional genes provide similar functions . Multiple mechanisms for such genetic redundancy have been recognized [42] , [43] , including homologous genes with overlapping functions , parallel metabolic or regulatory pathways , and other processes . Such buffering effects appear to be common since many genes are members of multi-gene families and many examples are known where two or more genes must be simultaneously inhibited to produce a phenotype . Finally , the phenotypic efficiency of an siRNA depends on many factors including the level of the targeted mRNA [44] and the half-life of its encoded protein . Notably , many proteins with half-lives in days cannot be sufficiently depleted within the time frame of an RNAi screen [39] . All of the above factors combine to mask phenotypes and reduce the number of genes that can be characterized through current RNAi screening . Our results provide useful approaches for conducting and interpreting genome-wide RNAi studies . Our statistical model provides a framework to describe a complicated biological assay , genome-wide RNAi , by modeling factors that can contribute to the final output . Such factors include general measurement errors that affect many assays , as well as factors specific to RNAi , such as off-target effects and gene accessibility . The model structure is not specific to influenza virus or even to virus infection . Thus we expect that the approach is applicable to genome wide RNAi studies of many cellular phenotypes . To test this , we applied the model-based estimation method to a collection of three genome-wide RNAi studies for HIV host factors ( Supplementary Text S2 , Section 5 ) . The resulting HIV-derived parameter estimates were comparable to those from influenza virus , although an even lower overlap among the implicated gene sets led to a higher estimate of the involvement rate , at 28% ( Supplementary Text S2 , Table S6 in Text S2 ) . As noted above , multiple findings imply that a key contributing factor to false-negative results is low gene accessibility . In future studies this could be addressed by performing screens in a variety of cell lines and possibly by expanding the time points and assay conditions used . In addition , insightful bioinformatics analysis with implicated genes could expand interpretation of host factors involved in virus infection by integrating information from other genome-wide or single gene studies . Insights from such analysis could be used to focus future studies on selected pathways to further test their involvement . Useful examples of such bioinformatic extensions include recent studies analyzing host genes required for HIV replication in the context of protein-protein interaction and other networking properties [45]–[47] . These studies offer experimentally useful insights into the functional organization of such genes and their interactions with HIV and , like the results presented here , also explicitly suggest that many host factors required for HIV replication remain to be discovered . Additional information will help to refine the working probability model and thus provide a more accurate description of genome-wide RNAi studies . Our model was necessarily based on available published data , which was generally limited to lists of genes implicated in the primary detection screen and confirmed in the secondary validation test . To enrich understanding and modeling such RNAi screens , we strongly support publishing full genome-wide data from such studies , including data on genes excluded either for cell toxicity or low expression . | Genome-wide RNA interference assays of gene functions offer the potential for systematic , global analysis of biological processes . A pressing challenge is to develop meta-analysis methods that effectively combine information from multiple studies . One puzzle is that implicated gene lists from independent studies of the same process often show relatively low overlap . This disagreement might arise from false-positive factors , such as imperfect gene targeting ( off-target effects ) , or from false negatives if separate studies access different components of large , complex systems . We present new methods to examine the relations between individual genome-wide RNAi studies , using studies of host genes in influenza virus replication as a test case . We find that cross-study agreement is greater than suggested by overlap of reported gene lists . This better agreement is evidenced by the strong relation of independent gene lists in functional pathways and protein interaction networks , and by a statistical model that relates multi-study , gene-level findings to factors driving correct , false-negative , and false-positive gene identification . Our analysis of multiple genome-wide studies predicts that there are many undetected host genes important for influenza virus infection , and that false negatives are the major concerns for genome-wide studies . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | Limited Agreement of Independent RNAi Screens for Virus-Required Host Genes Owes More to False-Negative than False-Positive Factors |
Zoonotic visceral leishmaniasis ( VL ) is a severe infectious disease caused by protozoan parasites of the genus Leishmania and the domestic dogs are the main urban parasite reservoir hosts . In Brazil , indirect fluorescence antibody tests ( IFAT ) and indirect enzyme linked immunosorbent assay ( ELISA ) using promastigote extracts are widely used in epidemiological surveys . However , their sensitivity and specificity have often been compromised by the use of complex mixtures of antigens , which reduces their accuracy allowing the maintenance of infected animals that favors transmission to humans . In this context , the use of combinations of defined peptides appears favorable . Therefore , they were tested by combinations of five peptides derived from the previously described Leishmania diagnostic antigens A2 , NH , LACK and K39 . Combinations of peptides derived A2 , NH , LACK and K39 antigens were used in ELISA with sera from 44 human patients and 106 dogs . Improved sensitivities and specificities , close to 100% , were obtained for both sera of patients and dogs . Moreover , high sensitivity and specificity were observed even for canine sera presenting low IFAT anti-Leishmania antibody titers or from asymptomatic animals . The use of combinations of B cell predicted synthetic peptides derived from antigens A2 , NH , LACK and K39 may provide an alternative for improved sensitivities and specificities for immunodiagnostic assays of VL .
Zoonotic visceral leishmaniasis ( VL ) caused by Leishmania infantum is an important emerging parasitic disease found in countries around the Mediterranean basin , in the Middle East , and in Latin America [1] , [2] . In these areas , wild canids constitute major sylvatic reservoirs , and domestic dogs are the principal urban reservoir hosts [3] , [4] . Hence , euthanasia of seropositive dogs has been adopted as a mainstay control measure in some countries [5] . However , domestic reservoir control programs may fail because of the high incidence of canine infection , the insensitivity of the diagnostic tests to detect infectious dogs and time delays between diagnosis and euthanasia by public health services [4] . Although adopted in European countries , treatment of infected dogs is not allowed in Brazil , based on the assumption that treated dogs may also remain as a source of parasites for sand fly infection . In this context , sensitive diagnostic tests , applicable to field conditions , are becoming increasingly necessary to facilitate and improve the control of disease [6] . Enzyme-linked immunosorbent assays ( ELISAs ) [7] and indirect fluorescence antibody tests ( IFAT ) [8] are widely used for serological diagnosis of VL . However , these tests present relative low sensitivity and specificity , which underestimates the actual rate of infection and allows the maintenance of infected animals and transmission . Several defined Leishmania antigens have been tested to overcome these difficulties and to improve both sensitivity and specificity [9] . Immunochromatographic tests for the diagnosis of leishmaniasis using the rK39 antigen has been evaluated in several countries , with variable results [6] , [10] , [11] . Development of effective diagnosis is also critical for control and possible eradication of visceral leishmaniasis and sensitive and specific rapid tests may be especially helpful to achieve this goal [12] . Therefore , there are still much room for improvement of serological diagnosis of VL , including identification and combination antigens and test formats . B cell epitopes prediction by bioinformatics analysis of protein sequences has been proposed as a good alternative to select peptides for diagnostic tests [13] , [14] . In the present study , we tested , in ELISA against sera from 44 patients and 106 dogs , combinations of predicted B cell peptides derived from A2 , NH , LACK and K39 , which have been previously evaluated as antigens for serodiagnosis of visceral leishmaniasis [15]–[21] . Improved sensitivity for detection of asymptomatic and symptomatic canine visceral leishmaniasis ( CVL ) , including canine sera with low anti-Leishmania antibody titers as detected by IFAT , and active disease in human patients was demonstrated for the majority of the peptide combinations .
Sera of dogs were obtained from already-existing collections ( Sera collection of the Laboratory of Molecular Biology of the Faculty of Pharmacy , Federal University of Minas Gerais ) . Approval to use the samples was obtained from institutional review board ( IRB ) - Comitê de Ética em Experimentação Animal ( CETEA ) from Universidade Federal de Minas Gerais ( UFMG ) , under the protocol 20/2010 . Sera of human patients were also obtained from an already-existing collection ( Sera collection of the Laboratory of Immunoparasitology of the Research Center René Rachou , Fundação Oswaldo Cruz ) . IRB approval to use the samples was obtained from the Institutional Committee on Ethics of Human Research of Fundação Oswaldo Cruz , under the protocol 12/2006 . All samples were analyzed anonymously . The aminoacid sequence of A2 ( amastigote stage-specific S antigen homolog of L . donovani ) , k39 ( kinesin related protein of L . chagasi ) , LACK ( Leishmania analogue of the receptor kinase C ) and NH ( nucleoside hydrolase ) proteins were subjected to analysis with software available online at http://www . expasy . org/tools/protscale . html . The analyses generate numerical and graphical scores to predict the position of linear B-cell epitopes . Peptides that fulfilled , at least in part , the criteria of high hydrophilicity ( Hoop &Woods ) , high alpha-helix structures ( Chou & Fasman ) , low coil ( Deleage & Roux ) , low beta-sheet structures ( Chou & Fasman ) , high percentage of accessible residues and low beta-turn structures ( Chou & Fasman ) were selected for synthesis and screening . The selected peptides were also submitted to BepiPred software ( http://www . cbs . dtu . dk/services/BepiPred/ ) to predict the location of linear B-cell epitopes using a combination of a hidden Markov model and a propensity scale method [13] . BepiPred software scores peptides according to hydrophilicity values , secondary structures and the probability of a aminoacid is located in certain positions as compared to other mapped B cell epitopes . Peptides displaying scores higher than 0 . 35 may be , therefore , considered putative B cell epitopes . Peptides were synthesized according to a standard N-9-ethyloxycarbonyl ( Fmoc ) strategy on a PSSM8 multispecific peptide synthesizer ( Shimadzu , Kyoto , Japan ) by solid-phase synthesis and were purified by high performance liquid chromatography and confirmed with a Micromass Q-Tof Micro ( Micromass MS Technologies , Division of Waters , Milford , MA ) and the peptides obtained by this method were all C-terminal amides . A panel of 106 canine sera was used in the study . Serum samples were divided into three groups based on history of exposure and infection status . Group 1 contained negative control sera from 14 healthy blood donor pets of various ages and breeds ( previously classified as seronegative dogs after ELISA-based assays for the detection of antibodies against parasite-specific recombinant antigens rK39 , rK26 , and rA2 ) that attended a veterinary clinic in Minas Gerais , Brazil . Group 2 contained 30 serum samples from clinically symptomatic ( n = 17 ) and asymptomatic ( n = 13 ) dogs in which L . infantum visceral infection was proven by the demonstration of the presence of the parasites in bone marrow specimens and/or necropsy tissue samples as previously reported [21] . All infected dogs enrolled in this group were selected during a longitudinal epidemiological survey of CVL carried out in a rural area of endemicity ( Pancas , ES; 2003–2004 ) in southeast Brazil [22] . Group 3 contained sera from 62 dogs with L . infantum infection from CVL endemic areas in Brazil . They had been previously tested in IFAT and ELISA . Sera presenting IFAT titers >1∶40 dilutions and ELISA optical densities>cut off values ( cut off values were determined by the mean of OD of 14 negative canine control sera plus two standard deviations ) were considered positive ( IgG ) for CVL . All 62 samples had their status confirmed by parasitological analyses which included the search for parasites in bone marrow aspirates by PCR , microscopic examination of Giemsa stained smears and culturing in NNN/LIT medium at 23°C , as previously described [23] . Since a significant correlation was observed between IFAT and ELISA tests for all sera samples ( data not shown ) , positive sera in group 3 were further grouped , according to their previous reactivity in IFAT , regardless its clinical status , as low ( n = 20 ) ( <1∶320 dilutions ) , intermediate ( n = 20 ) ( >1∶320 <1∶640 ) and high ( n = 22 ) ( >1∶640 ) IFAT titers . Human VL sera were obtained from patients with active visceral leishmaniasis ( n = 28 ) . Diagnosis of VL was defined when , besides clinical and epidemiologic features , amastigotes were seen at Giemsa stained smears of bone marrow aspirates or promastigote forms were identified on culture of peripheral blood or bone marrow aspirates . In the presence of suggestive clinical and epidemiologic characteristics , negative parasitological findings , but positive anti-Leishmania antibodies by IFAT or ELISA , definitive diagnostic was firmed after successful specific treatment . Control sera ( n = 16 ) were obtained from individuals living in Vale do Jequitinhonha ( in cities: São Pedro do Jequitinhonha , Caju , Virgem das Graças e Melquíades ) , a rural region of Minas Gerais State in southeast Brazil . None of the individuals presented signs of visceral leishmaniasis at clinical examination . All of them had negative results for specific Leishmania PCR in sera samples . Sera samples were also submitted to ELISA with the crude extract of the parasite to confirm that they were negative . Levels of total IgG immunoglobulin were measured by ELISA . Briefly , 96-well flexible PVC plates ( BD Biosciences , San Jose , CA ) were sensitized with 5 µg/mL of each synthetic peptide diluted in water ( 100 µL per well ) . The sensitized plate was left in the oven until dry and then was left overnight at 4°C . Plates were blocked with PBS-2% casein at 37°C for 1 h and treated successively with 1∶200 dilutions of canine serum samples for 1 h at 37°C . Peroxidase labeled antibodies specific to canine IgG ( Sigma , St . Louis , MO ) were diluted at 1∶5000 and added for 1 h at 37°C . Following another washing step , the enzyme bound to the immunosorbent is assayed by the addition of a chromogenic substrate , 3 , 3′ , 5 , 5′-tetramethylbenzidine ( TMB ) in citrate buffer containing hydrogen peroxide . Reactions were stopped by the addition of H2SO4 2N . Optical densities were determined at 450 nm in ELISA reader ( BioRad , Hercules , CA ) . Each sera sample was assayed in triplicate . The lower limit of positivity ( cut off ) was determined by the mean of OD of 14 negative canine control sera plus two standard deviations . Sensitization of the plates followed the same as described above . However , plates were sensitized with 40 µg/mL of each synthetic peptide diluted in water ( 100 µL per well , resulting in 4 µg/well ) . When 2 peptides were tested simultaneously ( peptides 13 and 47 , 13 and 19 , 18 and 19 , 17 and 47 and 19 and 47 ) , plates were sensitized with 20 µg/mL of each synthetic peptide . After antigen sensitization , the plates were blocked with 2% BSA at 37°C for 2 h and treated successively with 1∶100 dilutions of patients serum samples for 1 h at 37°C . After washing step , biotinilated labeled antibodies human-IgG ( Sigma , St . Louis , MO ) were diluted at 1∶5000 and added to the plate for 1 h at 37°C . Then , add the streptavidin-peroxidase conjugate diluted 1∶1000 for 30 minutes at 37°C . After three washes , the substrate 3 , 3′ , 5 , 5′-tetramethylbenzidine ( TMB ) in citrate buffer containing hydrogen peroxide were added to the plate . Reactions were stopped by the addition of H2SO4 2N . Optical densities were read at 450 nm in ELISA reader ( BioRad , Hercules , CA ) . Each sera sample was assayed in triplicate . The lower limit of positivity ( cut off ) was determined by the mean of OD of 16 negative human control sera plus two standard deviations . One-Way ANOVA test was used to compare the performances of the assays . A p value of less than 0 . 05 was considered significant . Sensitivity and specificity were calculated by binary classification test . The sensitivity and specificity for each test were calculated by using the formulas: Sensitivity = True positive/ ( True positive+False negative ) ×100% and Specificity = True negative/ ( True negative+False positive ) ×100% .
The study of the structure of proteins ( ProtScale ) allowed the selection of five peptides , as shown in Table 1: TPAVQKRVKEVGTKP and TTVVGNQTLEKVT , corresponding to numbers 17 and 18 peptides , respectively , derived from Nucleoside Hydrolase antigen , VVSTSRDGTAISWK , corresponding to peptide 19 , derived from LACK protein and ESTTAAKMSAEQDRESTRATLE , corresponding to peptide number 13 , derived from K39 protein . Additionally , in Table 1 is represented the peptide derived from the A2 protein , corresponding to peptide 47 ( VGPQSVGPLSVGPQSVGPLS ) . However , the inclusion of this peptide was based on previous analysis of epitope prediction and reactivity with sera of BALB/c mice vaccinated with the A2 antigen [24] . Among the peptides select , 4 ( peptides 13 , 17 , 19 and 47 ) also showed linear B-cell epitopes with significant values in the analysis by BepiPred , with emphasis on peptides 13 and 47 , derived from K39 and A2 antigens , which presented higher scores . By contrast , the peptide 18 showed a low score for the presence of linear B-cell epitopes . Initially , to test the reactivity of the selected peptides with antibodies present in dogs' sera with VL , the spot synthesis technique , followed by immunoassay was applied ( data not shown ) . The membranes were incubated with a pool of sera from infected and control animals . Four of the five peptides showed high intensity reactions with sera of dogs with confirmed VL as compared with the control group . Peptide 19 presented reactivity with both positive and negative sera . However , considering the use of pooled sera on the spot synthesis experiment , peptide 19 was also included in a more discriminatory analysis through ELISA with individual sera samples . All peptides were then tested against Group 2 sera , which included samples collected from symptomatic and asymptomatic animals ( Figure 1 ) . As shown in figure 1 , all peptides were able to detect as positives all the sera samples from both asymptomatic ( n = 13 ) and symptomatic animals ( n = 17 ) Figure 1 . No significant differences were observed in sensitivity between the two groups . The reactivity of the 5 peptides was further evaluated with a larger panel of sera from parasitological positive ( Group 3 ) or control dogs ( Group 1 ) , which included sera samples classified according to IFAT reactivity as low ( <1∶320 dilutions ) , intermediate ( >1∶320 <1∶640 ) and high ( >1∶640 ) titers ( Figure 2 ) . High sensitivity ( 100% ) was observed for all peptides to detect infection in dogs with high IFAT antibody titers ( 22 animals ) , when the peptides were tested individually ( Figure 2/Table 2 ) . However , decreased sensitivities ( varying between 55% and 90% ) were observed for all peptides when tested against sera of dogs with IFAT intermediate ( >1∶320 <1∶640 ) ( 20 animals ) or low antibody titers ( 1∶80 >1∶320 ) ( 20 animals ) ( Table 2 ) . Concerning specificity , a value of 100% was observed for peptide 17 and 92% for the other peptides , as shown in Table 2 . Since decreased sensitivity was detected for each peptide individually with sera of dogs with intermediate and low IFAT titers , we tested the hypothesis that the sensitivity to detect VL would increase by combining peptides in the same reaction . Assuming this would increase sensitivity by broaden instead of simply increasing the number of available epitopes for reaction , half of the concentration was used for each peptide instead of double the total peptide concentration . Combinations of two peptides were then tested against dog sera with low and intermediate antibody titers ( Figure 3/Table 2 ) . For sera with low antibody titers , the best results , i . e , improved sensitivities as compared to peptides tested individually , ranged from 90% to 95% , were obtained with combinations between the peptides 13 and 47 , 13 and 19 , 18 and 19 , 47 and 17 and 47 and 18 . Specificity was also improved , reaching 100% for all of these combinations . The sensitivity of peptide 47 was not altered when combined with peptide 17 , which in contrast , had its sensitivity improved from 85% to 90% ( Figure 3/Table 2 ) . For sera with intermediate antibody titers , improved sensitivities for both peptides were observed for combinations between peptides 13 and 19 , 18 and 19 , 47 and 18 , varying from 80% to 95% . The sensitivity of peptide 47 ( 85% ) was not affected when combined with peptide 13 and decreased when associated to peptide 17 ( Figure 3/Table 2 ) . Peptides were also tested ( individually and combined ) against human sera ( n = 44 ) including patients with active visceral leishmaniasis ( n = 28 ) and healthy individuals with previous negative results in ELISA to Leishmania ( n = 16 ) . The results obtained are shown in Figure 4 and 5/Table 3 . Sensitivity values of 82% , 93% and 96% were observed for peptides 18 , 47 and 13 , respectively . And ELISA with peptides 17 and 19 gave the best results , displaying sensitivities of 100% . The specificity for the peptides tested individually ranged from 81% to 94% . The combination of peptides brought an improvement in sensitivity and specificity between peptides 13 and 19 , 18 and 19 , 47 and 13 , and 19 and 47 where we observe a sensitivity of 100% for the first three combinations and 95 . 65% for the last one , respectively . Only for the combination between peptides 17 and 47 a reduction in sensitivity to 70 . 83% was observed . Specificity values for all combination ranged from 93 . 10% to 100% ( Figure 4 and 5/Table 3 ) .
In the present work , the Leishmania proteins A2 , K39 , LACK and NH were submitted to B cell epitope prediction and the derived synthetic peptides were evaluated through ELISA against sera of dogs and patients for the serodiagnosis of VL . Using the Protscale software , six different parameters were evaluated for each protein to select peptides . Considering the scores for these parameters , an adequate profile was observed for the majority of peptides , as compared to the minimal and maximum scores for the corresponding proteins , except for peptide 47 . Peptide 47 displayed the lower values for hydrophilicity and presence of alpha helix , which are expected to be high for B cell epitopes , and the highest values for coil and beta turn structures , which in contrast are expected to be low . On the other hand , prediction using BepiPred resulted in scores higher than 0 . 35 for all peptides , except for peptide 18 . Altogether , our results indicate that the two analyses may be complementary to each other and that this strategy is useful for selecting diagnostic antigens . Accurate diagnosis of canine leishmaniasis is essential towards a more efficient control of this zoonosis , but it remains problematic due to the high incidence of asymptomatic infections [25] . Initially , we tested the five peptides with sera from dogs clinically classified as asymptomatic and symptomatic ( Group 2 ) . It is noteworthy that the sera samples of Group 2 have been previously tested in ELISA using SLA or the recombinant proteins rA2 , rK39 and rK26 [21] . The retrospective analysis of the data obtained by Porrozzi et al . ( 2007 ) revealed that 4 , 9 , 5 and 4 out of 13 sera from asymptomatic animals included in the present study , were not reactive with SLA , rA2 , rK39 and rK26 , respectively , whereas all the symptomatic samples were positive when either rK26 , rK39 or SLA were used as antigens . Using rA2 , 6 symptomatic sera samples were identified as negative . Moreover , the majority of samples that were not reactive with these antigens were obtained from asymptomatic animals presenting low antibody titers in IFAT ( ≤1∶80 ) . In the present analysis , for both asymptomatic and symptomatic VL canine sera , sensitivities and specificities of 90% and 100% , respectively , were observed . Therefore , improved sensitivity was observed for assays using the synthetic peptides as compared to SLA and the recombinant proteins , especially for sera of asymptomatic animals . In this sense , our results largely confirm and improve the potential of these antigens for serodiagnosis of leishmaniasis . Detection of infection in animals with low or intermediated anti-Leishmania antibody titers , regardless their clinical status , is critical for diagnosis and control of VL . The failure to detect infection in these animals may contribute to the maintenance of parasite's transmission for both canine and human populations , one of the major factors that hinder control strategies . On the other hand , highly sensitive diagnosis may require combined antigens . As indicated by immunoproteomic approaches , Leishmania parasites display extensive variability in antigenic composition , and apparently absence of immunodominant antigens when individual sera samples are analyzed [26] , suggesting that single antigen diagnostic tests may display decreased sensitivities . Indeed , rK39 based tests may lack sensitivity for canine sera with low antibody concentration [19] . By combining two peptides , increased sensitivities ( 90–95% ) and specificity ( 100% ) were observed for dog sera with low IFAT antibodies titers . Similar findings were also observed for sera with intermediated antibodies titers ( 80–95% of sensitivity and 100% specificity ) . Improved sensitivities may have resulted from increased number of reactive epitopes , leading to increased OD readings and numbers of positive sera as compared to the reactivity with a single peptide . These findings will be particularly useful for diagnosis of dogs with low and intermediate titers of antibodies , since most current tests fail in this task . Since early and sensitive diagnosis is seen as a critical aspect for management and , possibly , eradication of human visceral leishmaniasis [12] , [27]–[29] , we have also investigated the reactivity of the peptides with sera of patients with active VL . Similarly , improved results have been observed when the combinations of peptides were tested against sera of human patients with active disease , suggesting that the epitopes selected were also recognized individually by human sera and that their serological reactivity may be independent and complementary , leading to an additive effect . Therefore , the association of peptides is an alternative to broaden the epitopes to be detected by antibodies , improving sensitivity . On the other hand , the absence of improved sensitivity for association between the peptides 17 and 47 may be explained by the presence of low levels of Leishmania specific antibodies in the control negative sera , since healthy controls were selected from endemic area and previous exposure of these individuals to parasite antigens may not be completely ruled out . In many endemic areas , VL frequently overlaps with the occurrence of other forms of leishmaniasis or even with other infectious diseases , such as tuberculosis and leprosy . Cross-reactivity with antibodies raised against other infectious diseases consists in an additional shortcoming for development of specific visceral leishmaniasis diagnosis . Cross-reactivity of synthetic peptides with sera of patients presenting other infections was not assessed in the present work . Therefore , additional investigations are further warranted to better determine peptides specificity . In conclusion , the combination of synthetic peptides , identified through B cell epitope predicition , may be useful for the development of highly sensitive and specific serodiagnosis for VL . The peptides identified may be especially interesting for the development sensitive immunochromatographic tests . Since these test format do not require sophisticated laboratory facilities or trained personnel staff to be routinely performed , and antibody quantification is not required for diagnosis of VL , they are more practical and easily applied , allowing rapid diagnosis in field conditions in endemic areas of difficult access to laboratory facilities [11] , [12] , [30]–[32] . Therefore , these peptides coupled to immunochromatographic tests may allow sensitive and early detection of infected dogs and their fast withdraw from transmission areas , regardless their antibody levels and clinical status , improving the control of VL in endemic areas . | Visceral leishmaniasis is endemic in many areas of tropical and subtropical America where it constitutes a significant public health problem . It is usually diagnosed by enzyme-linked immunosorbent assays ( ELISA ) using crude Leishmania antigens , but a variety of other immunological methods may also be applied . Although these approaches are useful , historically their sensitivity and specificity have often been compromised by the use of complex mixtures of antigens . In this context , the use of combinations of purified , well-characterized antigens appears preferable and may yield better results . In the present study , combinations of peptides derived from the previously described Leishmania diagnostic antigens A2 , NH , LACK and K39 were used in ELISA against sera from 106 dogs and 44 human patients . Improved sensitivities and specificities , close to 100% , for both sera of patients and dogs was observed for ELISA using some combinations of the peptides , including the detection of VL in dogs with low anti-Leishmania antibody titers and asymptomatic infection . So , the use of combinations of B cell predicted synthetic peptides derived from antigens A2 , NH , LACK and K39 may provide an alternative for improved sensitivities and specificities for immunodiagnostic assays of VL . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"biology",
"veterinary",
"science"
] | 2012 | Improved Canine and Human Visceral Leishmaniasis Immunodiagnosis Using Combinations of Synthetic Peptides in Enzyme-Linked Immunosorbent Assay |
Listeria monocytogenes is an opportunistic Gram-positive bacterial pathogen responsible for listeriosis , a human foodborne disease . Its cell wall is densely decorated with wall teichoic acids ( WTAs ) , a class of anionic glycopolymers that play key roles in bacterial physiology , including protection against the activity of antimicrobial peptides ( AMPs ) . In other Gram-positive pathogens , WTA modification by amine-containing groups such as D-alanine was largely correlated with resistance to AMPs . However , in L . monocytogenes , where WTA modification is achieved solely via glycosylation , WTA-associated mechanisms of AMP resistance were unknown . Here , we show that the L-rhamnosylation of L . monocytogenes WTAs relies not only on the rmlACBD locus , which encodes the biosynthetic pathway for L-rhamnose , but also on rmlT encoding a putative rhamnosyltransferase . We demonstrate that this WTA tailoring mechanism promotes resistance to AMPs , unveiling a novel link between WTA glycosylation and bacterial resistance to host defense peptides . Using in vitro binding assays , fluorescence-based techniques and electron microscopy , we show that the presence of L-rhamnosylated WTAs at the surface of L . monocytogenes delays the crossing of the cell wall by AMPs and postpones their contact with the listerial membrane . We propose that WTA L-rhamnosylation promotes L . monocytogenes survival by decreasing the cell wall permeability to AMPs , thus hindering their access and detrimental interaction with the plasma membrane . Strikingly , we reveal a key contribution of WTA L-rhamnosylation for L . monocytogenes virulence in a mouse model of infection .
Listeria monocytogenes ( Lm ) is a ubiquitous Gram-positive bacterium and the causative agent of listeriosis , a human foodborne disease with high incidence and morbidity in immunocompromised hosts and other risk groups , such as pregnant women , neonates and the elderly . Clinical manifestations range from febrile gastroenteritis to septicemia , meningitis and encephalitis , as well as fetal infections that can result in abortion or postnatal health complications [1] . The most invasive and severe forms of the disease are a consequence of the ability of this pathogen to overcome important physiological barriers ( intestinal epithelium , blood-brain barrier and placenta ) by triggering its internalization and promoting its intracellular survival into phagocytic and non-phagocytic cells . Once inside a host cell , a tightly coordinated life cycle , whose progression is mediated by several specialized bacterial factors , enables Lm to proliferate and spread to neighboring cells and tissues [2 , 3] . The Lm cell wall is composed of a thick peptidoglycan multilayer that serves as a scaffold for the anchoring of proteins , among which are several virulence factors [4] , and of glycopolymers such as teichoic acids , which account for up to 70% of the protein-free cell wall mass [5 , 6] . These anionic polymers are divided into membrane-anchored teichoic acids ( lipoteichoic acids , LTAs ) and peptidoglycan-attached teichoic acids ( wall teichoic acids , WTAs ) . In Listeria , WTAs are mainly composed of repeated ribitol-phosphate subunits , whose hydroxyl groups can be substituted with a diversity of monosaccharides [5] . While the polymer structure and the chemical identity of the substituent groups of LTAs are rather conserved across listeriae [7 , 8] , they display a high variability in WTAs , even within the same species [9] . Specific WTA substitution patterns are characteristic of particular Lm serotypes: N-acetylglucosamine is common to serogroups 1/2 and 3 , and to serotype 4b , but serogroup 1/2 also contains l-rhamnose , whereas serotype 4b displays d-glucose and d-galactose [10] . The broad structural and chemical similarity of LTAs and WTAs results in a considerable degree of functional redundancy , which has complicated the characterization of these macromolecules and the assignment of specific biological roles . However , studies on Gram-positive bacteria have revealed their contribution to important physiological functions ( e . g . cell envelope cationic homeostasis [11] , regulation of autolysin activity [12] , assembly of cell elongation and division machineries [13] , defense against antimicrobial peptides [14] ) and to virulence-promoting processes , such as adhesion and colonization of host tissues [15 , 16] . Antimicrobial peptides ( AMPs ) are a large family of small peptides ( <10 kDa ) produced by all forms of living organisms [17] , which constitute a major player of the innate immune response against microbial pathogens . Despite their structural diversity , the majority of AMPs share both cationic and amphipathic properties that favor respectively their interaction with the negatively charged prokaryotic surface and insertion into the plasma membrane [17 , 18] . Subsequent pore formation or other AMP-mediated membrane-disrupting mechanisms induce bacterial death through direct cell lysis or deleterious interaction with intracellular targets [19] . Bacteria have evolved multiple strategies to avert killing by AMPs [20 , 21] . One strategy consists in the modification of their cell surface charge , a process achieved mainly by masking anionic glycopolymers with positively charged groups , thus decreasing their affinity to AMPs . In Gram-positive pathogens , d-alanylation of teichoic acids is a well-characterized mechanism and was demonstrated to be important for bacterial resistance to host-secreted AMPs [22 , 23] . In contrast , the contribution of WTA glycosylation mechanisms in AMP resistance has not yet been investigated . We have previously reported genome-wide transcriptional changes occurring in Lm strain EGD-e during mouse infection [24] . Our analysis revealed an elevated in vivo expression of the lmo1081-1084 genes , here renamed as rmlACBD because of the high homology of the corresponding proteins with enzymes of the l-rhamnose biosynthesis pathway . In this work , we show that the decoration of Lm WTAs with l-rhamnose requires the expression of not only the rmlACBD locus but also of rmlT , an upstream-flanking gene encoding a putative rhamnosyltransferase . We also demonstrate that Lm becomes more susceptible to AMPs in the absence of WTA l-rhamnosylation and predict that this effect is due to an increase of the Lm cell wall permeability to these bactericides , which results in a faster disruption of the plasma membrane integrity with lethal consequences for the bacterial cell . Importantly , we present evidence that this WTA tailoring process is required for full-scale Lm virulence in the mouse model of infection .
To identify new Lm genes potentially critical for the infectious process , we previously performed the first in vivo transcriptional profiling of Lm EGD-e . Among the Lm genes displaying the largest increase in transcription throughout infection , we identified a set of previously uncharacterized genes that are included in a pentacistronic operon ( lmo1080 to lmo1084 ) [25] . This operon is found in L . monocytogenes strains belonging to serogroups 1/2 , 3 and 7 , and is absent from serogroup 4 strains [26] ( Fig 1 ) . Interestingly , aside from Listeria seeligeri 1/2b strains , this locus is not found in any other Listeria spp . , such as the nonpathogenic Listeria innocua or the ruminant pathogen Listeria ivanovii , which pinpoints it as a genetic feature of a particular subset of pathogenic Listeria strains and suggests that its expression may be important to Listeria pathogenesis in humans . The four proteins encoded by the lmo1081-lmo1084 genes share a high amino acid sequence homology with the products of the rmlABCD gene cluster . These genes are widely distributed among Gram-negative ( e . g . Salmonella enterica [27] , Shigella flexneri [28] , Vibrio cholerae [29] , Pseudomonas aeruginosa [30] ) and Gram-positive species ( e . g . Mycobacterium tuberculosis [31] , Streptococcus mutans [32] , Geobacillus tepidamans [33] , Lactobacillus rhamnosus [34] ) ( Fig 1 ) , the majority of which being known pathogens or potentially pathogenic . Despite the inter-species variability observed in the genetic organization of the rml genes , the respective proteins exhibit a remarkable degree of conservation ( S1 Table in S1 Text ) . In light of this , we renamed the lmo1081-lmo1084 genes to rmlACBD , respectively ( Fig 1 ) . The RmlABCD proteins catalyze the conversion of glucose-1-phosphate to a thymidine-diphosphate ( dTDP ) -linked form of l-rhamnose [35] ( S1A Fig in S1 Text ) , which is a component of the WTAs from most Listeria strains possessing the rml genes [6] . To address the role of rmlACBD in Lm WTA glycosylation with l-rhamnose , we constructed an Lm EGD-e derivative mutant strain lacking the rmlACBD locus ( ΔrmlACBD ) ( S2A Fig in S1 Text ) and investigated if the absence of these genes could affect the WTA l-rhamnosylation status . We prepared WTA hydrolysates from exponential phase cultures of wild type ( EGD-e ) , ΔrmlACBD and a complemented ΔrmlACBD strain expressing rmlACBD from its native promoter within an integrative plasmid ( ΔrmlACBD+rmlACBD ) . Samples were resolved by native PAGE and the gel stained with Alcian blue to visualize WTA polymer species . A mutant strain unable to synthesize WTAs ( ΔtagO1ΔtagO2 ) [36] was used to confirm that the detected signal corresponds to WTAs . Compared to the wild type sample , the ΔrmlACBD WTAs displayed a shift in migration , which was reverted to a wild type-like profile in WTAs from the ΔrmlACBD+rmlACBD sample ( Fig 2A ) , indicating that the native WTA composition requires the presence of the rmlACBD genes . To confirm this , we investigated the WTA carbohydrate composition from these strains . WTA polymers were isolated from cell walls purified from bacteria in exponential growth phase , hydrolyzed and analyzed by high-performance anion exchange chromatography coupled with pulsed amperometric detection ( HPAEC-PAD ) to detect monosaccharide species . WTA extracts obtained from ΔrmlACBD bacteria completely lacked l-rhamnose , in contrast to those isolated from the parental wild type strain ( Fig 2B ) . The role of rmlACBD in Lm WTA l-rhamnosylation was definitely confirmed by the analysis of WTAs from ΔrmlACBD+rmlACBD bacteria , in which l-rhamnose was detected at levels similar to those observed in the wild type sample ( Fig 2B ) . Similar observations were made with purified cell wall samples that contain WTAs still attached to the peptidoglycan matrix ( S3A Fig in S1 Text ) . The absence of muramic acid , one of the peptidoglycan building blocks , from WTA extracts ( Fig 2B ) indicates that l-rhamnose is specifically associated with WTAs and is not a putative peptidoglycan contaminant . This is corroborated by the absence of l-rhamnose in purified peptidoglycan samples ( Fig 2C ) . WTAs have been identified as important regulators of peptidoglycan cross-linking and maturation [37] . To investigate if l-rhamnose decoration of WTAs has any involvement in the maturation of the Lm peptidoglycan , we performed HPLC analysis of the muropeptide composition of mutanolysin-digested peptidoglycan samples from wild type , ΔrmlACBD and ΔrmlACBD+rmlACBD bacteria . No differences in the nature and relative amount of muropeptide species were observed between strains ( S3B Fig in S1 Text ) , ruling out a role for WTA l-rhamnosylation in the consolidation of the peptidoglycan architecture . Overall , these results confirm that a functional rmlACBD locus is required for the association of l-rhamnose with Lm WTAs , likely by providing the molecular machinery responsible for the synthesis of l-rhamnose . The rml operon in Lm includes a fifth gene , lmo1080 , located upstream of rmlA ( Fig 1 ) , which codes for a protein similar to the B . subtilis minor teichoic acid biosynthesis protein GgaB , shown to possess sugar transferase activity [38] . Conserved domain analysis of the translated Lmo1080 amino acid sequence revealed that its N-terminal region is highly similar ( e-value 10–22 ) to a GT-A family glycosyltransferase domain ( S1B Fig in S1 Text ) . In GT-A enzymes , this domain forms a pocket that accommodates the nucleotide donor substrate for the glycosyl transfer reaction , and contains a signature DxD motif necessary to coordinate a catalytic divalent cation [39] . This motif is also found within the predicted glycosyltransferase domain sequence of Lmo1080 as a DHD tripeptide ( S1B Fig in S1 Text ) . For these reasons , we investigated whether Lmo1080 , which we renamed here RmlT ( for l-rhamnose transferase ) , was involved in the l-rhamnosylation of Lm WTAs . We constructed an Lm EGD-e mutant strain lacking rmlT ( S2A Fig in S1 Text ) and analyzed the structure and sugar composition of its WTAs as described above . WTAs isolated from ΔrmlT bacteria displayed a faster migration in gel ( Fig 2A ) and did not contain any trace of l-rhamnose ( Fig 2B ) , fully recapitulating the ΔrmlACBD phenotype . Reintroduction of a wild type copy of rmlT into the mutant strain ( ΔrmlT+rmlT ) resulted in a phenotype that resembles that of the wild type strain , with regards to WTA gel migration profile ( Fig 2A ) and presence of l-rhamnose in the WTA fraction ( Fig 2B ) . To discard the possibility that the deletion of rmlT exerted a negative polar effect on the downstream expression of rmlACBD , potentially disrupting the synthesis of l-rhamnose used for WTA glycosylation , we compared the transcription of the rmlACBD genes in the wild type and ΔrmlT Lm strains by quantitative real-time PCR . Transcript levels were unchanged in the ΔrmlT background as compared to the wild type strain ( S2B Fig in S1 Text ) , indicating that the deletion of rmlT did not interfere with the transcription of rmlACBD . To definitely confirm that Lm ΔrmlT still holds the capacity to synthesize l-rhamnose , being only incapable to incorporate it in nascent WTA polymers , we evaluated the presence of l-rhamnose in the cytoplasmic compartment of this strain . The intracellular content of early exponential-phase bacteria from the wild type , ΔrmlACBD and ΔrmlT strains was extracted , hydrolyzed and analyzed by HPAEC-PAD to compare the sugar composition of cytoplasmic extracts . As shown in Fig 2D , a peak corresponding to l-rhamnose was detected in the cytoplasmic samples from the wild type and ΔrmlT strains , but not from the ΔrmlACBD strain , clearly demonstrating that , as opposed to ΔrmlACBD bacteria , ΔrmlT bacteria retain a functional l-rhamnose biosynthesis pathway . These results indicate that the depletion of l-rhamnose observed in ΔrmlT WTAs is a consequence of the absence of the WTA l-rhamnosyltransferase activity performed by RmlT . Therefore , we propose RmlT as the glycosyltransferase in charge of decorating Lm WTAs with l-rhamnose . WTAs were previously associated with bacterial resistance against salt stress [40] and host defense effectors , such as lysozyme [37 , 41] . We thus investigated the potential involvement of WTA l-rhamnosylation in these processes by assessing the growth of the ΔrmlACBD and ΔrmlT strains in the presence of high concentrations of either NaCl or lysozyme . As shown in Fig 3A , no significant difference was observed between the growth of the wild type and the two mutant strains in BHI broth containing 5% NaCl . Similarly , no difference was detected between the growth behavior of these strains after the addition of different concentrations of lysozyme ( 50 μg/ml and 1 mg/ml ) to bacterial cultures in the exponential phase ( Fig 3B ) . As expected , we observed an immediate and significant decrease in the survival of the lysozyme-hypersensitive ΔpgdA mutant [42] ( Fig 3B ) . These data demonstrate that Lm does not require l-rhamnosylated WTAs to grow under conditions of high osmolarity nor to resist the cell wall-degrading activity of lysozyme . WTAs were also found to be involved in bacterial resistance to host-secreted defense peptides [14 , 43] . To investigate the role of WTA l-rhamnosylation in Lm resistance to AMPs , we evaluated the in vitro survival of wild type , ΔrmlACBD and ΔrmlT Lm , as well as of the respective complemented strains , in the presence of biologically active synthetic forms of AMPs produced by distinct organisms: gallidermin , a bacteriocin from the Gram-positive bacterium Staphylococcus gallinarum [44]; CRAMP , a mouse cathelicidin [45] , or its human homolog LL-37 [46] . After two hours of co-incubation with different AMP concentrations , surviving bacteria were enumerated by plating in solid media . The overall survival levels of Lm varied with each AMP , evidencing their distinct antimicrobial effectiveness ( S4 Fig in S1 Text ) . However , when compared to the wild type strain , the ΔrmlACBD and ΔrmlT mutants displayed a consistent decrease in their survival levels in the presence of any of the three AMPs ( Fig 3C ) , in a dose-dependent manner ( S4 Fig in S1 Text ) . Restoring WTA l-rhamnosylation through genetic complementation of the mutant strains resulted in an increase of the survival rate to wild type levels . This result demonstrated the important contribution of l-rhamnosylated WTAs towards Lm resistance against AMPs , pointing to a role for WTA glycosylation in bacterial immune evasion mechanisms . The increased AMP susceptibility of Lm strains defective in WTA l-rhamnosylation suggests that this process is required to hinder the bactericidal activity of AMPs . Since AMPs generally induce bacterial death by disrupting the integrity of the plasma membrane , we hypothesized that the higher susceptibility of the ΔrmlACBD and ΔrmlT mutant strains resulted from an increased AMP-mediated destabilization of the Lm membrane . In this context , two scenarios were envisioned: i ) AMPs could be binding with higher affinity to the l-rhamnose-deficient Lm cell wall , or ii ) they could be crossing it at a faster pace , thus reaching the membrane more quickly than in wild type Lm . To explore these possibilities , we first investigated the binding affinity of the mouse cathelicidin CRAMP towards Lm cell walls depleted of l-rhamnose . For this , we incubated the different Lm strains with CRAMP for a short period and analyzed by flow cytometry the amount of Lm-bound peptide exposed at the cell surface and accessible for antibody recognition . We detected fluorescence associated with surface-exposed CRAMP in all strains ( Fig 4A ) . However , the mean fluorescence intensity ( MFI ) values were significantly reduced in both ΔrmlACBD and ΔrmlT mutants , in comparison to wild type Lm and the complemented strains ( Fig 4A and 4B ) . This suggests that CRAMP was less accessible to immunolabeling at the cell surface of Lm lacking l-rhamnosylated WTAs . The affinity of AMPs towards the bacterial surface is driven by electrostatic forces between positively charged peptides and the anionic cell envelope [23] . To determine if variations of the Lm surface charge contributed to the reduced amount of CRAMP exposed at the surface of ΔrmlACBD and ΔrmlT bacteria , we compared the surface charge of Lm with or without l-rhamnosylated WTAs . For this , we analyzed the binding of cytochrome c , a small protein with positive charge at physiological conditions ( isoelectric point ~10 ) , to the wild type and mutant Lm strains . As positive control , we used a mutant strain that cannot modify its LTAs with d-alanine ( ΔdltA ) and , as a result , displays a higher surface electronegativity and a concomitant higher affinity for positively charged compounds [14 , 47] . As expected , the level of cytochrome c binding was higher with the ΔdltA strain than with the respective wild type strain , as illustrated by a decreased percentage of unbound cytochrome c ( Fig 4C ) . However , no significant difference in cytochrome c binding levels was observed between ΔrmlACBD , ΔrmlT and wild type EGD-e strains ( Fig 4C ) , indicating that the absence of l-rhamnose in WTAs does not affect the Lm surface charge . This was further corroborated by zeta potential measurements showing similar pH-dependent variations for both wild type and mutant strains ( S5 Fig in S1 Text ) . Overall , these results allowed us to discard electrostatic changes as a reason behind the difference in the levels of CRAMP detected at the Lm cell surface . To further explore the decreased levels of surface-exposed CRAMP in Lm strains lacking l-rhamnosylated WTAs , we compared total levels of bacterium-associated CRAMP in the different strains by flow cytometry , following a short incubation with a fluorescently labeled form of this AMP . The intensity of Lm-associated CRAMP fluorescence was comparable for the wild type EGD-e , ΔrmlACBD and ΔrmlT strains ( Fig 4D and 4E ) , indicating that the overall peptide levels associated to Lm cells were similar between the different strains . Accordingly , the residual fluorescence in the supernatants obtained by centrifugation of the bacteria-peptide suspensions was also similar ( Fig 4F ) . As positive control we used the ΔdltA strain , which displayed a significantly stronger peptide binding than its parental wild type strain ( Fig 4D–4F ) . These data strongly suggest that the increased CRAMP susceptibility of Lm strains lacking l-rhamnosylated WTAs results from an improved penetration of CRAMP through their cell walls . Altogether , these results showed that l-rhamnosylated WTAs do not interfere with the Lm surface charge or with the binding efficiency of AMPs , but likely promote Lm survival by hindering the crossing of its cell wall by these bactericidal molecules . In light of these results , we then examined whether WTA l-rhamnosylation interfered with the dynamics of AMP interaction with the Lm plasma membrane . We performed a time-course study to follow Lm membrane potential changes induced by CRAMP . In live bacteria , the membrane potential is an electric potential generated across the plasma membrane by the concentration gradients of sodium , potassium and chloride ions . Physical or chemical disruption of the plasma membrane integrity leads to the suppression of this potential ( depolarization ) [48] . Lm strains were incubated with DiOC2 ( 3 ) , a green fluorescent voltage-sensitive dye that readily enters into bacterial cells . As the intracellular dye concentration increases with higher membrane potential , it favors the formation of dye aggregates that shift the fluorescence emission to red . After stabilization of the DiOC2 ( 3 ) fluorescence , CRAMP was added to bacterial samples and the rate of Lm depolarization was immediately analyzed by measuring the red fluorescence emission decline in a flow cytometer . The decrease in the membrane potential was consistently greater in the ΔrmlACBD and ΔrmlT strains as compared to wild type Lm , particularly in the first 10–15 min ( Fig 5A ) , indicating that the Lm plasma membrane integrity is compromised faster by the action of CRAMP in the absence of l-rhamnosylated WTAs . To investigate if increased CRAMP-mediated disruption of the Lm membrane integrity was associated with increased permeabilization , we monitored in real time the entry of the fluorescent probe SYTOX Green into the different Lm strains , following the addition of CRAMP . This probe only enters into bacterial cells with a compromised membrane and displays a strong green fluorescence emission after binding to nucleic acids . As expected , when CRAMP was omitted from the bacterial suspensions , any increase in SYTOX Green-associated fluorescence was detected ( Fig 5B ) . However , in the presence of the peptide , the green fluorescence intensity of samples containing the ΔrmlACBD or ΔrmlT mutants increased earlier than in samples containing wild type Lm ( Fig 5B ) , eventually reaching similar steady-state levels at later time points ( S7 Fig in S1 Text ) . These observations indicate that the CRAMP-mediated permeability increase of the Lm membrane to SYTOX Green occurs faster in strains lacking l-rhamnosylated WTAs . To investigate the ultrastructural localization of the peptide , we performed immunoelectron microscopy on CRAMP-treated wild type and ΔrmlACBD Lm strains . Interestingly , CRAMP-specific labeling was not only detected in the Lm cell envelope , as expected , but also in the cytoplasm ( Fig 5C ) , suggesting that this AMP may additionally target components or processes inside Lm . Comparison of the subcellular distribution of CRAMP between these two bacterial compartments revealed a preferential cell envelope localization in wild type Lm , which contrasted with the slight but significantly higher cytoplasmic localization of the peptide in the ΔrmlACBD strain ( Fig 5D ) . These observations are in agreement with a model in which CRAMP crosses the Lm cell wall more efficiently in the absence of WTA l-rhamnosylation , therefore reaching the bacterial membrane and the cytoplasm comparatively faster . Finally , to confirm that the presence of l-rhamnosylated WTAs hinders the capacity of AMPs to flow through the Lm cell wall , we assessed levels of CRAMP retained in purified cell wall samples from the wild type , ΔrmlACBD and ΔrmlT strains by Western blot . After incubation with CRAMP , peptides trapped within the peptidoglycan matrix were released by mutanolysin treatment of the cell wall and quantitatively resolved by SDS-PAGE . Immunoblotting revealed a small but consistent decrease in the amount of peptide associated with the cell wall from the two mutant strains in comparison with wild type Lm ( Fig 5E and 5F ) . This result indicates that the lack of l-rhamnose in WTAs results in a partial loss of the AMP retention capacity of the Lm cell wall , which induces an enhanced AMP targeting of the Lm plasma membrane and consequent bacterial killing . All combined , these data support a model where the l-rhamnosylation of WTAs alters the Lm cell wall permeability to favor the entrapment of AMPs . This obstructive effect hinders AMP progression through the cell wall and delays their lethal interaction with the plasma membrane . To evaluate the importance of WTA l-rhamnosylation in Lm pathogenicity , we assessed the in vivo virulence of Lm strains lacking l-rhamnosylated WTAs . BALB/c mice were inoculated orally with wild type , ΔrmlACBD or ΔrmlT strains , and the bacterial load in the spleen and liver of each animal was quantified three days later . The proliferative capacity of both ΔrmlACBD and ΔrmlT mutant strains was similarly reduced in both organs , although more significantly in the liver ( Fig 6A and 6B ) . To determine if the decreased virulence of the mutant strains was due to a specific defect in the crossing of the intestinal epithelium , BALB/c mice were challenged intravenously , bypassing the intestinal barrier . Three days post-infection , the differences between mutant and wild type strains , in both organs , were similar to those observed in orally infected animals ( Fig 6C and 6D ) , thus discarding any sieving effect of the intestinal epithelium on the decreased splenic and hepatic colonization by both ΔrmlACBD and ΔrmlT . Importantly , organs of mice infected intravenously with the complemented strains ( ΔrmlACBD+rmlACBD and ΔrmlT+rmlT ) displayed bacterial loads comparable to wild type Lm-infected organs ( Fig 6C and 6D ) . The attenuated in vivo phenotype of the ΔrmlACBD and ΔrmlT strains was not caused by an intrinsic growth defect , as demonstrated by their wild type-like growth profiles in broth or inside eukaryotic cells ( S8 Fig in S1 Text ) . These results confirmed the involvement of the rml operon in virulence , revealing a significant contribution of WTA l-rhamnosylation to Lm pathogenesis . Importantly , the in vivo attenuation of the ΔrmlT strain , which is unable to append l-rhamnose to its WTAs but is able to synthesize the l-rhamnose precursor , showed that although l-rhamnose biosynthesis is required to achieve optimal levels of virulence it is its covalent linkage to the WTA backbone that is crucial for the successful Lm host infection . To evaluate the protective role of WTA l-rhamnosylation against AMPs in vivo , we performed virulence studies in a CRAMP-deficient mouse model . To determine the influence of WTA l-rhamnosylation in Lm intestinal persistence , we performed oral infections of adult CRAMP knockout 129/SvJ mice ( cramp-/- , KO ) [49] and of age- and background-matched wild type mice ( cramp+/+ , WT ) , with the wild type or ΔrmlACBD Lm strains and monitored the respective fecal carriage . In both WT and KO mice , we observed comparable dynamics of fecal shedding of the wild type and ΔrmlACBD strains ( Fig 6E and 6F ) . In agreement with the comparable virulence defects observed for WTA l-rhamnosylation-deficient bacteria , following oral or intravenous inoculation of BALB/c mice ( Fig 6A–6D ) , these results suggest a minor role for CRAMP in the control of Lm during the intestinal phase of the infection . We then inoculated WT and KO mice intravenously and quantified bacterial numbers in the spleen and liver , three days post-infection . In line with what was observed in BALB/c mice ( Fig 6C ) , the ΔrmlACBD strain showed significant virulence attenuation in both organs of WT mice ( Fig 6G ) . Interestingly , this virulence defect was nearly abolished in KO animals , with the ΔrmlACBD strain displaying an organ-colonizing capacity similar to wild type bacteria ( Fig 6H ) . In addition , bacterial loads were higher in the organs of KO mice than in those of WT animals ( Fig 6G and 6H ) . These data indicate that , in comparison to their WT congeners , KO mice are more susceptible to Lm infection , and confirm the in vivo listericidal activity of CRAMP . Altogether , these results highlight a key role for host-produced CRAMP in restraining Lm infection and demonstrate that WTA l-rhamnosylation also promotes resistance to AMPs in an in vivo context .
Teichoic acids are key players in the maintenance of the Gram-positive cell envelope integrity and functionality . They are typically decorated with d-alanine and/or a variety of glycosyl groups , which influence the overall properties of these polymers [9] . Whereas d-alanylation of WTAs has been demonstrated to contribute towards bacterial defense against AMPs [14 , 23] , the involvement of glycosylation in this process has never been investigated . In this study , we show for the first time that the glycosylation of Lm WTAs with l-rhamnose is mediated by the WTA l-rhamnosyltransferase RmlT and confers protection against AMPs in vitro and during mouse infection . Based on our data , we propose that this protection results from a delayed traversal of the Lm cell envelope by AMPs in the presence of l-rhamnose-decorated WTAs . Most importantly , we reveal a key role for l-rhamnosylated WTAs in the processes underlying Lm pathogenesis . Unlike S . aureus or B . subtilis [22] , WTAs in Listeria are not decorated with d-alanine , undergoing only glycosylation with a small pool of monosaccharides [6 , 10] . Among these is l-rhamnose , which is the product of a remarkably conserved biosynthetic pathway that is encoded by the rmlABCD genes [35] . Interestingly , a significant number of bacteria harboring these genes are commonly pathogenic [27–32] and have l-rhamnose in close association with surface components [50 , 51] . In Listeria , the rmlACBD locus is only found in certain serotypes of Lm ( 1/2a , 1/2b , 1/2c , 3c and 7 ) and L . seeligeri ( 1/2b ) . These serotypes were all shown to have l-rhamnose in their WTAs , except for Lm serotypes 3c and 7 [6] , which appear to be unable to produce this sugar because of mutations within rmlA and rmlB , respectively ( Fig 1 ) . Our results confirmed that the appendage of l-rhamnose to Lm WTAs requires the products of the rmlACBD locus . Ultimately , WTA glycosylation is catalyzed by glycosyltransferases , a class of enzymes that recognize nucleotide-sugar substrates and transfer the glycosyl moiety to a WTA subunit [52] . In silico analysis of lmo1080 , the first gene of the operon including rmlACBD ( Fig 1 ) showed that it encodes a protein with putative glycosyltransferase activity . The genomic location and predicted protein function were strong indicators that this gene might encode the transferase involved in the l-rhamnosylation of Lm WTAs . Our data demonstrated that whereas lmo1080 , that we renamed rmlT , is dispensable for rhamnose biosynthesis , it is required for the addition of l-rhamnose to WTAs in Lm strains with a functional l-rhamnose pathway , thus validating RmlT as the l-rhamnose-specific WTA glycosyltransferase in Lm . WTAs are associated with the natural resistance of S . aureus to peptidoglycan-degrading enzymes , such as lysozyme [37 , 41] . In contrast , absence of WTA decoration , but not of the polymers , was shown to induce an increase of the staphylococcal susceptibility to lysostaphin [53] . Modifications of the Lm peptidoglycan , such as N-deacetylation [42] , were found to contribute to protection against lysozyme , but the role of WTAs and in particular their decoration , was never addressed . Our results discard WTA l-rhamnosylation as a component of the Lm resistance mechanism to this host immune defense protein , as well as its involvement in the promotion of growth under osmotic conditions . Other innate immune effectors , such as antimicrobial peptides ( AMPs ) , also target bacterial organisms [54] that in turn have developed resistance strategies to avoid injury and killing induced by AMPs . Among these strategies is the reshaping and fine-tuning of cell envelope components to lower AMP affinity to the bacterial surface [21] . Previous studies showed a clear link between the d-alanylation of WTAs and AMP resistance [14 , 43] . In this context , we found here a similar role for WTA l-rhamnosylation , showing that , in the absence of l-rhamnosylated WTAs , bacteria exhibit an increased susceptibility to AMPs produced by bacteria , mice and importantly by humans . Although from such distinct sources , AMPs used here share a cationic nature that supports their activity . However , while teichoic acid d-alanylation is known to reduce the cell wall electronegativity [14] , glycosyl substituents of Lm WTAs are neutrally charged and WTA glycosylation should thus promote AMP resistance through a different mechanism . It is well established that AMPs induce bacterial death mainly by tampering with the integrity of the plasma membrane . This can be achieved through multiple ways , all of which are driven by the intrinsic amphipathic properties of this class of peptides [55] . Nonetheless , the initial interaction of AMPs with bacterial surfaces is mediated by electrostatic forces between their positive net charge and the anionic cell envelope [23] . Our data show that , unlike d-alanylation [56] , WTA l-rhamnosylation does not interfere with the Lm cell surface charge , in agreement with l-rhamnose being an electrostatically neutral monosaccharide . Importantly , the reduced levels of surface-exposed CRAMP in Lm strains lacking l-rhamnosylated WTAs suggested instead that their increased susceptibility to this peptide was correlated with its improved penetration of the l-rhamnose-depleted Lm cell wall . We confirmed this premise with data showing that CRAMP-mediated cell depolarization and plasma membrane permeabilization events occur earlier in WTA l-rhamnosylation-deficient Lm strains . In addition , we also observed a predominant cytoplasmic presence of CRAMP in these mutant strains , in contrast to the preferential cell envelope localization in wild type Lm , further suggesting a WTA l-rhamnosylation-dependent kinetic discrepancy in the progression of CRAMP through the Lm cell envelope . Saar-Dover et al . demonstrated in the WTA-lacking Streptococcus agalactiae ( GBS ) that LTA d-alanylation promoted resistance to the human cathelicidin LL-37 by hindering cell wall crossing and plasma membrane disturbance [57] . They proposed that the underlying mechanism does not rely on modulation of the surface charge but on LTA conformation-associated alterations of the cell wall packing density [57] . Our data are in line with these observations and although we did not detect changes in the cell wall cross-linking status , we cannot ignore a possible impact of l-rhamnosylation on WTA polymer conformation accounting for changes in cell wall permeability . If one considers that the peptidoglycan , a multi-layered and compact structure , is densely populated with WTA polymers decorated with multiple units of the rather bulky l-rhamnose molecule , spatial constraints and increased cell wall density need to be accounted . In fact , we showed that purified Lm cell wall depleted of l-rhamnose does not retain CRAMP in its peptidoglycan matrix as effectively as cell wall containing l-rhamnosylated WTAs . In addition , we have indications that soluble l-rhamnose interferes with CRAMP activity , improving the survival of WTA l-rhamnosylation mutants of Lm . These observations suggest a potential interaction between l-rhamnose and AMPs , which could favor the “retardation effect” that ultimately promotes Lm survival . We previously reported a significantly increased transcription of rmlACBD during mouse spleen infection [24] , which suggested that WTA l-rhamnosylation is highly activated by Lm to successfully infect this host organ . Our infection studies in mice confirmed the importance of this mechanism for Lm pathogenesis by revealing a significant virulence attenuation of WTA l-rhamnosylation-deficient Lm strains . Surprisingly , the expression of rmlT appeared unchanged during mouse spleen infection as compared to growth in BHI [24] , suggesting that an increased L-rhamnose biosynthesis could be sufficient to induce an increased WTA l-rhamnosylation and AMP resistance . Faith et al . also observed a decreased bacterial burden of a serotype 4b Lm strain lacking the gtcA gene [58] , a mutation that resulted in complete loss of galactose decoration of its WTAs [59] . Interestingly , gtcA is also present in Lm EGD-e , where it appears to be involved in WTA substitution with N-acetylglucosamine [60] , and was shown to contribute to the colonization of the mouse spleen , liver and brain [61] . However the mechanism through which this occurs remains unclear . Virulence studies in mice lacking the CRAMP gene corroborated our in vitro susceptibility data and revealed the importance of WTA l-rhamnosylation-promoted resistance to AMPs for Listeria virulence . In vivo data also provided a strong insight into the protective role of CRAMP against systemic infection by Lm , as had been previously observed with other bacterial pathogens [49 , 62 , 63] . Our results on fecal shedding dynamics suggest that the contribution of CRAMP to the control of Lm during the intestinal phase of infection is minimal . A previous report showed a negligible enteric secretion of CRAMP in normal adult mice [64] , which may explain the similar shedding behavior of the wild type and ΔrmlACBD strains that were observed in both mouse strains . In this scenario , infection studies in newborn animals , whose enterocytes actively express CRAMP [45 , 64] , may provide conclusive information regarding the role of WTA l-rhamnosylation in the Lm resistance to CRAMP during the intestinal phase of the infection . Notwithstanding , CRAMP is actively produced by phagocytes in adult mice [65] . As a major target for Lm colonization , the spleen is also an important reservoir of phagocytic cells . We can speculate that WTA l-rhamnosylation is particularly important to increase the chances of Lm surviving CRAMP-mediated killing during spleen infection . Considering our data on the Lm susceptibility to LL-37 , the human homolog of CRAMP , we can also envisage this scenario in the context of human infection . In conclusion , our work has unveiled for the first time a role for WTA glycosylation in bacterial resistance to AMPs . We propose that WTA l-rhamnosylation reduces the cell wall permeability to AMPs , promoting a delay in the crossing of this barrier and in the disruption of the plasma membrane , thus favoring Lm survival and virulence in vivo . Our findings reveal a novel facet in the contribution of WTA modifications towards AMP resistance , reinforcing the crucial role of these Gram-positive surface glycopolymers in host defense evasion .
Bacterial strains used in this study are listed in Table 1 . Lm and E . coli strains were routinely cultured aerobically at 37°C in brain heart infusion ( BHI , Difco ) and Lysogeny Broth ( LB ) media , respectively , with shaking . For experiments involving the Lm ΔtagO1ΔtagO2 strain , bacteria were first cultured overnight at 30°C with shaking in the presence of 1 mM IPTG ( isopropyl-β-d-thiogalactopyranoside ) , washed and diluted ( 1:100 ) in fresh BHI and cultured overnight at 30°C with shaking [36] . When appropriate , the following antibiotics were included in culture media as selective agents: ampicilin ( Amp ) , 100 μg/ml; chloramphenicol ( Cm ) , 7 μg/ml ( Lm ) or 20 μg/ml ( E . coli ) ; erythromycin ( Ery ) , 5 μg/ml . For genetic complementation purposes , colistin sulfate ( Col ) and nalidixic acid ( Nax ) were used at 10 and 50 μg/ml , respectively . Lm mutant strains were constructed in the EGD-e background through a process of double homologous recombination mediated by the suicide plasmid pMAD [66] . DNA fragments corresponding to the 5’- and 3’-flanking regions of the rmlACBD locus ( lmo1081—4 ) were amplified by PCR from Lm EGD-e chromosomal DNA with primers 1–2 and 3–4 ( S2 Table in S1 Text ) , and cloned between the SalI—MluI and MluI—BglII sites of pMAD , yielding pDC303 . Similarly , DNA fragments corresponding to the 5’- and 3’-flanking regions of rmlT ( lmo1080 ) were amplified with primers 15–16 and 17–18 ( S2 Table in S1 Text ) , and cloned between the SalI—EcoRI and EcoRI—BglII sites of pMAD , yielding pDC491 . The plasmid constructs were introduced in Lm EGD-e by electroporation and transformants selected at 30°C in BHI—Ery . Positive clones were re-isolated in the same medium and grown overnight at 43°C . Integrant clones were inoculated in BHI broth and grown overnight at 30°C , after which the cultures were serially diluted , plated in BHI agar and incubated overnight at 37°C . Individual colonies were tested for growth in BHI—Ery at 30°C and antibiotic-sensitive clones were screened by PCR for deletion of rmlACBD ( primers 5–6 , 7–8 , 9–10 and 11–12 ) and rmlT ( primers 19–20 ) ( S2 Table in S1 Text ) . Genetic complementation of the deletion mutant strains was performed as described [24] . DNA fragments containing either the rmlACBD or rmlT loci were amplified from Lm EGD-e chromosomal DNA with primers 13–14 and 21–22 ( S2 Table in S1 Text ) , respectively , and cloned between the SalI—PstI sites of the phage-derived integrative plasmid pPL2 [67] , generating pDC313 and pDC550 . The plasmid constructs were introduced in the E . coli strain S17-1 and transferred , respectively , to the ΔrmlACBD and ΔrmlT strains by conjugation on BHI agar . Transconjugant clones were selected in BHI—Cm/Col/Nax and chromosomal integration of the plasmids confirmed by PCR with primers 23 and 24 ( S2 Table in S1 Text ) . All plasmid constructs and gene deletions were confirmed by DNA sequencing . Total bacterial RNA was isolated from 10 ml of exponential cultures ( OD600 = 0 . 6 ) by the phenol-chloroform extraction method , as previously described [68] , and treated with DNase I ( Turbo DNA-free , Ambion ) , as recommended by the manufacturer . Purified RNAs ( 1 μg ) were reverse-transcribed with random hexamers , using iScript cDNA Synthesis kit ( Bio-Rad Laboratories ) . Quantitative real-time PCR ( qPCR ) was performed in 20-μl reactions containing 2 μl of cDNA , 10 μl of SYBR Green Supermix ( Bio-Rad Laboratories ) and 0 . 25 μM of forward and reverse primers ( S2 Table in S1 Text ) , using the following cycling protocol: 1cycle at 95°C ( 3 min ) and 40 cycles at 95°C ( 30 s ) , 55°C ( 30 s ) and 72°C ( 30 s ) . Each target gene was analyzed in triplicate and blank ( water ) and DNA contamination controls ( unconverted DNase I-treated RNA ) were included for each primer pair . Amplification data were analyzed by the comparative threshold ( ΔΔCt ) method , after normalization of the test and control sample expression values to a housekeeping gene ( 16S rRNA ) . For qualitative analysis , PCR was performed in 20-μl reactions containing 2 μl of cDNA , 10 μl of MangoMix 2× reaction mix ( Bioline ) and 0 . 5 μM of forward and reverse qPCR primers , using the following protocol: 1 cycle at 95°C ( 5 min ) , 25 cycles at 95°C ( 30 s ) , 55°C ( 30 s ) and 72°C ( 20 s ) , and 1 cycle at 72°C ( 5 min ) . Amplification products were resolved in 1% ( w/v ) agarose gel and analyzed in a GelDoc XR+ System ( Bio-Rad Laboratories ) . Extraction and analysis of Lm WTAs by polyacrylamide gel electrophoresis was performed essentially as described [69] , with the exception that WTAs extracts were obtained from exponential-phase cultures . Sedimented bacteria were washed ( buffer 1: 50 mM MES buffer , pH 6 . 5 ) and boiled for 1 h ( buffer 2: 4% SDS in buffer 1 ) . After centrifugation , the pellet was serially washed with buffer 2 , buffer 3 ( 2% NaCl in buffer 1 ) and buffer 1 , before treatment with 20 μg/ml proteinase K ( 20 mM Tris-HCl , pH 8; 0 . 5% SDS ) at 50°C for 4 h . The digested samples were thoroughly washed with buffer 3 and distilled water and incubated overnight ( 16 h ) with 0 . 1 M NaOH , under vigorous agitation . Cell wall debris were removed by centrifugation ( 10 , 000 rpm , 10 min ) and the hydrolyzed WTAs present in the supernatant were directly analyzed by native PAGE in a Tris-tricine buffer system . WTA extracts were resolved through a vertical ( 20 cm ) polyacrylamide ( 20% ) gel at 20 mA for 18 h ( 4°C ) . To visualize WTAs , the gel was stained in 0 . 1% Alcian blue ( 40% ethanol; 5% acetic acid ) for 30 min and washed ( 40% ethanol; 10% acetic acid ) until the background is fully cleared . Optionally , for increased contrasting , silver staining can be performed on top of the Alcian blue staining . Cell walls of Lm strains were purified as described before [70] , with modifications . Overnight cultures were subcultured into 1–2 liters of BHI broth ( initial OD600 = 0 . 005 ) and bacteria grown until exponential phase ( OD600 = 1 . 0–1 . 5 ) . Cultures were rapidly cooled in an ice/ethanol bath and bacteria harvested by centrifugation ( 7 , 500 rpm , 15 min , 4°C ) . The pellet was resuspended in cold ultrapure water and boiled for 30 min with 4% SDS to kill bacteria and inactivate cell wall-modifying enzymes . The samples were cleared of SDS by successive cycles of centrifugation ( 12 , 000 rpm , 10 min ) and washing with warm ultrapure water until no detergent was detected [71] . SDS-free samples were resuspended in 2 ml of ultrapure water and cell walls disrupted with glass beads in a homogenizer ( FastPrep , Thermo Savant ) . Fully broken cell walls were separated from glass beads by filtration ( glass filters , pore size: 16–40 μm ) and from unbroken cell walls and other debris by low-speed centrifugation ( 2 , 000 rpm , 15 min ) . Nucleic acids were degraded after incubation ( 2 h ) at 37°C with DNase ( 10 μg/ml ) and RNase ( 50 μg/ml ) in a buffer containing 50 mM Tris-HCl , pH 7 . 0 , and 20 mM MgSO4 . Proteins were then digested overnight at 37°C with trypsin ( 100 μg/ml ) in the presence of 10 mM CaCl2 . Nuclease and proteases were inactivated by boiling in 1% SDS , and samples were centrifuged ( 17 , 000 rpm , 15 min ) and washed twice with ultrapure water . Cell walls were resuspended and incubated ( 37°C , 15 min ) in 8 M LiCl and then in 100 mM EDTA , pH 7 . 0 , after which they were washed twice with water . After resuspension in acetone and sonication ( 15 min ) , cell walls were washed and resuspended in ultrapure water before undergoing lyophilization . To obtain purified peptidoglycan , cell walls ( 20 mg ) were incubated for 48 h with 4 ml of 46% hydrofluoric acid ( HF ) , under agitation at 4°C . Samples were washed with 100 mM Tris-HCl , pH 7 . 0 , and centrifuged ( 17 , 000 rpm , 30 min , 4°C ) as many times as necessary to neutralize the pH . The pellet was finally washed twice with water prior to lyophilization . WTA extracts were obtained by incubating 1 mg of cell wall with 300 μl of 46% HF ( 18 h , 4°C ) . After centrifugation ( 13 , 200 rpm , 15 min , 4°C ) , the supernatant was recovered and evaporated under a stream of compressed air . The dried WTA residue was resuspended in water and lyophilized . The intracellular content of Lm strains was isolated according to a modified version of the protocol by Ornelas-Soares et al . [72] . Bacterial cultures ( 200 ml ) were grown until early exponential phase ( OD600 = 0 . 3 ) , and vancomycin was added at 7 . 5 μg/ml ( 5×MIC value [73] ) to induce the cytoplasmic accumulation of the peptidoglycan precursor UDP-MurNAc-pentapeptide . Cultures were grown for another 45 min and chilled in an ice-ethanol bath for 10 min . Bacteria were then harvested by centrifugation ( 12 , 000 rpm , 10 min , 4°C ) , washed with cold 0 . 9% NaCl , resuspended in 5 ml of cold 5% trichloroacetic acid and incubated for 30 min on ice . Cells and other debris were separated by centrifugation ( 4 , 000 rpm , 15 min , 4°C ) and the supernatant was extracted with 1–2 volumes of diethyl ether as many times as necessary to remove TCA ( sample pH should rise to at least 6 . 0 ) . The aqueous fraction containing the cytoplasmic material was lyophilized and the dried residue resuspended in ultrapure water . To analyze their sugar composition , purified cell wall and peptidoglycan ( 200 μg each ) , as well as cytoplasmic ( 500 μg ) and WTA extracts were hydrolyzed in 3 M HCl for 2 h at 95°C . After vacuum evaporation , the samples were washed with water and lyophilized . The hydrolyzed material was then resuspended in 150 μl of water and resolved by high-performance anion-exchange chromatography coupled with pulsed amperometric detection ( HPAEC-PAD ) . Ten microliters were injected into a CarboPac PA10 column ( Dionex , Thermo Fisher Scientific ) and eluted at 1 ml/min ( 30°C ) with 18 mM NaOH , followed by a gradient of NaCH3COO: 0–20 mM ( t = 25–30 min ) , 20–80 mM ( t = 30–35 min ) , 80–0 mM ( t = 40–45 min ) . Standards for glucosamine , muramic acid , l-rhamnose and ribitol ( Sigma-Aldrich ) were eluted under the same conditions to enable identification of chromatogram peaks . Data were acquired and analyzed with the Chromeleon software ( Dionex , Thermo Fisher Scientific ) . Muropeptide samples were prepared and analyzed as described [74] , with minor changes . Purified peptidoglycan was digested with 200 μg/ml mutanolysin ( Sigma-Aldrich ) in 12 . 5 mM sodium phosphate , pH 5 . 5 , for 16 h at 37°C . Enzymatic activity was halted by heating at 100°C for 5 min , after which the digested sample was reduced for 2 h with 2 . 5 mg/ml of sodium borohydride ( NaBH4 ) in 0 . 25 M borate buffer , pH 9 . 0 . The reaction was stopped by lowering the sample pH to 2 with ortho-phosphoric acid . After centrifugation , the supernatant was analyzed by reverse phase HPLC . Fifty microliters were injected into a Hypersil ODS ( C18 ) column ( Thermo Fisher Scientific ) and muropeptide species eluted ( 0 . 5 ml/min , 52°C ) in 0 . 1 M sodium phosphate , pH 2 . 0 , with a gradient of 5–30% methanol and detected at 206 nm . Mouse macrophage-like J774A . 1 cells ( ATCC , TIB-67 ) were propagated in Dulbecco’s modified Eagle’s medium ( DMEM ) containing 10% fetal bovine serum and infection assays were performed as described [24] . Briefly , cells ( ~2×105/well ) were infected for 45 min with exponential-phase bacteria at a multiplicity of infection of ~10 and treated afterwards with 20 μg/ml gentamicin for 75 min . At several time-points post-infection , cells were washed with PBS and lysed in cold 0 . 2% Triton X-100 for quantification of viable intracellular bacteria in BHI agar . One experiment was performed with triplicates for each strain and time-point . Lm cultures grown overnight were appropriately diluted in BHI broth and their growth under the presence of stressful stimuli was monitored by optical density measurement at 600 nm ( OD600 ) . For comparative analysis of Lm resistance to salt stress , bacterial cultures were diluted 100-fold in BHI alone ( control ) or BHI containing 5% NaCl . To assess the Lm resistance to lysozyme , exponential-phase cultures ( OD600 ≈ 1 . 0 ) were challenged with different doses of chicken egg white lysozyme ( Sigma ) . A mutant Lm strain hypersensitive to lysozyme ( ΔpgdA ) was used as a positive control for susceptibility . Bacteria in the exponential phase of growth ( OD600 = 0 . 7–0 . 8 ) were diluted ( 104 CFU/ml ) in sterile PB medium ( 10 mM phosphate buffer , pH 7 . 4; 1% BHI ) and mixed in a 96-well microplate with increasing concentrations of gallidermin ( Santa Cruz Biotechnology ) , CRAMP or LL-37 ( AnaSpec ) . Bacterial suspensions without AMPs were used as reference controls for optimal growth/survival . After incubation for 2 h at 37°C , the mixtures were serially diluted in sterile PBS and plated in BHI agar for quantification of viable bacteria . Each condition was analyzed in duplicate in three independent assays . Cytochrome c binding assays were performed as described [56] . Bacteria from mid-exponential-phase cultures ( OD600 = 0 . 6–0 . 7 ) were washed in 20 mM MOPS buffer , pH 7 . 0 , and resuspended in ½ volume of 0 . 5 mg/ml equine cytochrome c ( Sigma-Aldrich ) in 20 mM MOPS buffer , pH 7 . 0 . After 10 min of incubation , bacteria were pelleted and the supernatant collected for quantification of the absorbance at 530 nm . The mean absorbance values from replicate samples containing bacteria were subtracted to the mean value of a reference sample lacking bacteria , and the results were presented for each strain as percentage of unbound cytochrome c . Bacteria ( 1 ml ) from mid-exponential-phase cultures were washed twice with deionized water and diluted ( 107 CFU/ml ) in 15 mM NaCl solutions adjusted to different pH values ( 1 to 7 ) with nitric acid . Bacterial suspensions ( 750 μl ) were injected into a disposable capillary cell cuvette ( DTS1061 , Malvern Instruments ) and the zeta potential was measured at 37°C in a ZetaSizer Nano ZS ( Malvern Instruments ) , under an automated field voltage . Samples were measured in triplicate in three independent assays . Bacteria from 500 μl of mid-exponential-phase cultures were washed twice with PBS and treated for 5 min with 5 μg/ml CRAMP or PBS ( untreated control ) . After centrifugation , the supernatant was removed and PBS-washed bacteria were incubated for 1 h with rabbit anti-CRAMP ( 1:100 , Innovagen ) , followed by 1 h with Alexa Fluor 488-conjugated anti-rabbit IgG ( 1:200 , Molecular Probes ) . Finally , bacteria were fixed with 3% paraformaldehyde for 15 min , washed and resuspended in PBS . Alternatively , bacteria were similarly treated with an N-terminally 5-FAM-labeled synthetic form of CRAMP ( 95% purity , Innovagen ) , washed and resuspended in PBS . Samples were acquired in a FACSCalibur flow cytometer equipped with CellQuest software ( BD Biosciences ) and data were analyzed with FlowJo ( TreeStar Inc . ) . Green fluorescence was collected from at least 50 , 000 FSC/SSC-gated bacterial events in the FL1 channel ( 530 nm/20 nm bandpass filter ) . Fluorescence intensities were plotted in single-parameter histograms and results were presented as the average mean fluorescence intensity ( MFI ) value from three independent analyses . For bacterial membrane potential studies , the lipophilic fluorescent probe DiOC2 ( 3 ) ( 3 , 3-diethyloxacarbocyanine , Santa Cruz Biotechnology ) was used as a membrane potential indicator [48 , 75] . Mid-logarithmic phase bacteria were diluted ( 106 CFU/ml ) in PBS with 30 μM DiOC2 ( 3 ) and incubated for 15 min in the dark . CRAMP was added to a final concentration of 50 μg/ml and the sample was immediately injected in the flow cytometer . Control samples treated with PBS or with 1 . 5 mM sodium azide ( uncoupling agent ) were analyzed to determine the fluorescence values corresponding to basal ( 100% ) and null ( 0% ) membrane potential ( S6 Fig in S1 Text ) . Green and red ( FL3 , 670 nm/long bandpass filter ) fluorescence emissions were continuously collected from FSC/SSC-gated bacteria for 30 min . After acquisition , a ratio of red over green fluorescence ( R/G ) was calculated per event and plotted in the y-axis versus time . A series of consecutive one-minute-wide gates was applied to the plot and the mean R/G value per gate was determined . The mean R/G values from uncoupler-treated samples were deducted from the corresponding values from the untreated and CRAMP-treated samples , and the resulting values for each condition were normalized as percentage of the initial value ( t = 1 min ) . Finally , the temporal variation of the Lm membrane potential was represented graphically as the ratio of the normalized values from CRAMP-treated over untreated samples . Bacterial uptake of the cell-impermeable SYTOX Green dye was used to study membrane permeabilization induced by CRAMP [57] . Exponential-phase bacteria were washed and resuspended ( 107 CFU/ml ) in sterile PBS containing 1 μM SYTOX Green ( Molecular Probes ) . After 20 min of incubation in the dark , bacterial suspensions were mixed in PCR microplate wells with 50 μg/ml CRAMP or PBS ( negative control ) for a total volume of 100 μl . The mixtures were immediately placed at 37°C in a real-time PCR detection system ( iQ5 , Bio-Rad Laboratories ) and fluorescence emission at 530 nm was recorded every minute following excitation at 488 nm . One-hundred micrograms of purified cell wall were resuspended in 50 μl of 5 μg/ml CRAMP or PBS ( negative control ) and gently shaken for 5 min . Samples were centrifuged ( 16 , 000 × g , 1 min ) , washed in PBS and in TM buffer ( 10 mM Tris-HCl , 10 mM MgCl2 , pH 7 . 4 ) before overnight incubation at 37°C with mutanolysin ( 400 U/ml ) in TM buffer ( 50 μl ) . Supernatants were resolved by tricine-SDS-PAGE in a 16% gel , transferred onto nitrocellulose membrane and blotted with rabbit anti-CRAMP ( 1:1000 ) or mouse anti-InlA ( L7 . 7; 1:1000 ) , followed by HRP-conjugated goat anti-rabbit or anti-mouse IgG ( 1:2000 , P . A . R . I . S ) . Immunolabeled bands were visualized using SuperSignal West Dura Extended Duration Substrate ( Pierce ) and digitally acquired in a ChemiDoc XRS+ system ( Bio-Rad Laboratories ) . Exponential-phase bacteria treated with 50 μg/ml CRAMP for 15 min at 37°C were fixed for 1 h at room temperature ( 4% paraformaldehyde , 2 . 5% glutaraldehyde , 0 . 1 M sodium cacodylate , pH 7 . 2 ) , stained with 1% osmium tetroxide for 2 h and resuspended in 30% BSA ( high-purity grade ) . Bacterial pellets obtained after centrifugation in microhematocrit tubes were fixed overnight in 1% glutaraldehyde , dehydrated in increasing ethanol concentrations , and embedded in Epon 812 . Ultrathin sections ( 40–50 nm ) were placed on 400-mesh Formvar-coated copper grids and treated with 4% sodium metaperiodate and 1% periodic acid ( 10 min each ) for antigen retrieval . For immunogold labeling of CRAMP , sections were blocked for 10 min with 1% BSA and incubated overnight ( 4°C ) with rabbit anti-CRAMP ( 1:100 in 1% BSA ) . After extensive washing , sections were labeled with 10-nm gold complex-conjugated anti-rabbit IgG ( 1:200 in 1% BSA ) for 2 h , washed and contrasted with 4% uranyl acetate and 1% lead citrate . Images were acquired in a Jeol JEM-1400 transmission electron microscope equipped with a Gatan Orius SC1000 CCD camera and analyzed using ImageJ software . Virulence studies were done in mouse models of the following strains: wild type BALB/c and 129/SvJ ( Charles River Laboratories ) ; and CRAMP-deficient ( cramp-/- ) 129/SvJ , which was bred in our facilities from a breeding pair provided by Dr . Richard L . Gallo ( University of California , USA ) [49] . Infections were performed in six-to-eight week-old specific-pathogen-free females as described [76] . Briefly , for oral infections , 12-h starved animals were inoculated by gavage with 109 CFU in PBS containing 150 mg/ml CaCO3 , while intravenous infections were performed through the tail vein with 104 CFU in PBS . In both cases , the infection was carried out for 72 h , at which point the animals were euthanatized by general anesthesia . The spleen and liver were aseptically collected , homogenized in sterile PBS , and serial dilutions of the organ homogenates plated in BHI agar . For analysis of Lm fecal carriage , total feces produced by each infected animal ( n = 5 per strain ) up to a given time-point were collected , homogenized in PBS and serial dilutions plated in Listeria selective media ( Oxoid ) for bacterial enumeration . Mice were maintained at the IBMC animal facilities , in high efficiency particulate air ( HEPA ) filter-bearing cages under 12 h light cycles , and were given sterile chow and autoclaved water ad libitum . All the animal procedures were in agreement with the guidelines of the European Commission for the handling of laboratory animals ( directive 2010/63/EU ) , with the Portuguese legislation for the use of animals for scientific purposes ( Decreto-Lei 113/2013 ) , and were approved by the IBMC Animal Ethics Committee , as well as by the Direcção Geral de Veterinária , the Portuguese authority for animal protection , under license PTDC/SAU-MIC/111581/2009 . Statistical analyses were performed with Prism 6 ( GraphPad Software ) . Unpaired two-tailed Student’s t-test was used to compare the means of two groups; one-way ANOVA was used with Tukey’s post-hoc test for pairwise comparison of means from more than two groups , or with Dunnett’s post-hoc test for comparison of means relative to the mean of a control group . Mean differences were considered statistically non-significant ( ns ) when p value was above 0 . 05 . For statistically significant differences: * , p≤0 . 05; ** , p≤0 . 01; *** , p≤0 . 001 . | Listeria monocytogenes is a foodborne bacterial pathogen that preferentially infects immunocompromised hosts , eliciting a severe and often lethal disease . In humans , clinical manifestations range from asymptomatic intestinal carriage and gastroenteritis to harsher systemic states of the disease such as sepsis , meningitis or encephalitis , and fetal infections . The surface of L . monocytogenes is decorated with wall teichoic acids ( WTAs ) , a class of carbohydrate-based polymers that contributes to cell surface-related events with implications in physiological processes , such as bacterial division or resistance to antimicrobial peptides ( AMPs ) . The addition of other molecules to the backbone of WTAs modulates their chemical properties and consequently their functionality . In this context , we studied the role of WTA tailoring mechanisms in L . monocytogenes , whose WTAs are strictly decorated with monosaccharides . For the first time , we link WTA glycosylation with AMP resistance by showing that the decoration of L . monocytogenes WTAs with l-rhamnose confers resistance to host defense peptides . We suggest that this resistance is based on changes in the permeability of the cell wall that delay its crossing by AMPs and therefore promote the protection of the bacterial membrane integrity . Importantly , we also demonstrate the significance of this WTA modification in L . monocytogenes virulence . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | L-Rhamnosylation of Listeria monocytogenes Wall Teichoic Acids Promotes Resistance to Antimicrobial Peptides by Delaying Interaction with the Membrane |
Insulin is a major regulator of metabolism in metazoans , including the fruit fly Drosophila melanogaster . Genome-wide association studies ( GWAS ) suggest a genetic basis for reductions of both insulin sensitivity and insulin secretion , phenotypes commonly observed in humans with type 2 diabetes mellitus ( T2DM ) . To identify molecular functions of genes linked to T2DM risk , we developed a genetic tool to measure insulin-like peptide 2 ( Ilp2 ) levels in Drosophila , a model organism with superb experimental genetics . Our system permitted sensitive quantification of circulating Ilp2 , including measures of Ilp2 dynamics during fasting and re-feeding , and demonstration of adaptive Ilp2 secretion in response to insulin receptor haploinsufficiency . Tissue specific dissection of this reduced insulin signaling phenotype revealed a critical role for insulin signaling in specific peripheral tissues . Knockdown of the Drosophila orthologues of human T2DM risk genes , including GLIS3 and BCL11A , revealed roles of these Drosophila genes in Ilp2 production or secretion . Discovery of Drosophila mechanisms and regulators controlling in vivo insulin dynamics should accelerate functional dissection of diabetes genetics .
Insulin is a major regulator of metabolism , growth and development in metazoans , including the fruit fly Drosophila melanogaster . Insulin resistance in the liver and other human tissues can lead to compensatory increases in insulin production and secretion by pancreatic β cells , a facultative response that fails during pathogenesis of type 2 diabetes mellitus ( T2DM ) [1] . The decline of both insulin sensitivity and insulin secretion may have a genetic basis in humans [2] . Drosophila could emerge as a powerful system for dissecting the genetics of insulin resistance and secretion if appropriate physiological assays , like quantification of circulating insulin , could be used to assess insulin dynamics . Drosophila Insulin-like peptide 2 , 3 and 5 ( Ilp2 , 3 , and 5 ) are synthesized and secreted by insulin producing cells ( IPCs ) , median neurosecretory cells located in the pars intercerebralis , and are crucial hormonal regulators of development , growth and metabolism [3] , [4] . Ilp2 is a principal circulating insulin in flies , and is essential for maintaining normoglycemia [5] . Structural and biochemical studies of Drosophila insulin-like peptide association with its receptor suggest that Ilps might circulate at picomolar levels , similar to mammals [6] . Other than in mammals , however , no methods for determining the absolute concentration of circulating insulin with picomolar sensitivity exist , to our knowledge . The most widely-used method for assessing insulin secretion by Drosophila IPCs involves estimating the immuno-reactive signal for Ilp2 in IPCs [7] . By comparing the relative intensity of signal between experimental and control conditions , increased signal has been interpreted to indicate reduced or impaired secretion of insulin from IPCs . However , in using intracellular Ilp2 immunoreactivity as a surrogate for secretion , this method does not differentiate between changes in insulin production and secretion . To overcome these challenges , focus has shifted to the use of enzyme-linked immunosorbent assay ( ELISA ) as a potential method for Ilp2 quantification . An immunoepitope tagged Ilp2 was used to measure circulating Ilp2 by ELISA [8] , but the tagged Ilp2 was overexpressed in IPCs , making it difficult to assess physiological regulation of Ilp2 production and secretion . A recent study used polyclonal antibodies to measure circulating Ilp2 and Ilp5 in adult hemolymph by ELISA [9] , but only relative changes were reported . Moreover , the specialized nature of required reagents , like synthetic Ilp standards , have limited widespread adoption of this assay . Polypeptide-based immunoepitope tags can facilitate ELISA construction , but prior attempts over several decades to epitope-tag insulin , which undergoes extensive post-translational modification [6] , have led invariably to loss or elimination of bioactivity [10] , [11] . Maintenance of bioactivity in an epitope-tagged insulin would ensure that native mechanisms controlling crucial elements of insulin biology , like processing , storage , secretion and clearance , are being assayed . However , the bioactivity of prior epitope-tagged forms of Ilp2 has not been demonstrated or quantified [12] . Here we report the successful labeling of Ilp2 , a crucial regulator of glucose metabolism in Drosophila , with two immuno-epitopes at specific positions that preserved Ilp2 bioactivity . Using unique fly strains expressing double-tagged Ilp2 ( Ilp2HF ) , we developed robust and sensitive ELISA methods for quantifying circulating Ilp2 at picomolar concentration in Drosophila , and show that only a small fraction of total Ilp2 is secreted from IPCs in vivo and in vitro , like in mammals . Our studies reveal changes in either Ilp2 expression or secretion resulting from IPC-specific knockdown of Drosophila cognates of human T2DM genome-wide association study ( GWAS ) candidate loci , demonstrating genetic and molecular mechanisms linking these risk genes to insulin regulation . In addition , we uncovered previously undetected forms of genetic insulin receptor haploinsufficiency accompanied by adaptive insulin hypersecretion . Tissue specific dissection of this phenotype revealed a critical role for insulin signaling in the fat body in feedback regulation of systemic insulin levels . Thus , we provide the community with a potent new Drosophila tool for studies of insulin biology , integrative physiology and the genetic basis of human metabolic diseases .
To measure a circulating Drosophila insulin directly , we sought to tag Drosophila Insulin-like peptide 2 ( Ilp2 ) with immuno-detectable epitopes while preserving its in vivo bioactivity . Broad misexpression of a transgene encoding Ilp2 from GAL4-responsive upstream activation sequences ( UAS ) is lethal [13] , providing an assay of Ilp2 activity in vivo . To screen for permissive epitope-insertion sites that preserved the bioactivity of Ilp2 , we misexpressed transgenes encoding variant forms of hemagglutinin- ( HA ) and FLAG-epitope-tagged Ilp2 ( Figure S1 ) , and scored the resulting lethality . Like mammalian insulins , Drosophila Ilp2 is comprised of ‘B-chain’ and ‘A-chain’ polypeptides linked by disulfide bonds ( Figure 1A ) . Systematic variation of epitope position in the B-chain and A-chain of Ilp2 led to identification of tagged forms that remained lethal when broadly expressed , including a variant with HA-epitope fused to the B-chain carboxy-terminus and FLAG-epitope fused to the A-chain amino terminus ( hereafter called “Ilp2HF”; Figure 1A ) . In contrast , a previously described FLAG-epitope labeled Ilp2 [12] achieved only 38% lethality ( Figure S1 ) . Substitution of a conserved A-chain cysteine by tyrosine ( from a missense mutation called ‘Akita’ ) impairs insulin processing and activity in rodents [14] , [15] , and the orthologous substitution ( C119Y ) prevented Ilp2HF-induced lethality ( Figure S1 ) . Thus , we identified epitope-tags and positions in Ilp2 that preserved in vivo bioactivity in a synthetic lethality screen . To assess if Ilp2HF also retained native Ilp2 function and activity in regulating development , growth , and metabolism , we generated a 2 . 4 kilobase pair genomic fragment ( Figure 1B ) containing the native Ilp2 or Ilp2HF gene under the control of the endogenous Ilp2 regulatory sequence ( gd2 and gd2HF , respectively ) . We next assessed if developmental , growth , and metabolic defects observed in flies lacking Ilp2 , Ilp3 , and Ilp5 ( Ilp2–3 , 5 ) [5] could be rescued by introducing gd2 or gd2HF into Ilp2–3 , 5 mutants . Development from egg to adult eclosion in Ilp2–3 , 5 mutant females requires an average of 16 days , compared to 10 days for control flies ( Figure S2 ) . The delay is shortened to 11 days in mutants harboring a genomic Ilp2 rescue construct ( Ilp2–3 , 5 gd2 ) and 12 days in mutants harboring a genomic Ilp2HF rescue construct ( Ilp2–3 , 5 gd2HF; Figure S2 ) . Thus , both gd2 and gd2HF substantially rescued the developmental delay observed in insulin deficient flies , although the eclosion time was delayed by 1–2 days with gd2HF rescue . In addition , both gd2 and gd2HF rescued the reduced wing length of Ilp2–3 , 5 mutant adult flies to the same degree ( Figure 1C ) . In Ilp2–3 , 5 mutant flies , elevated levels of trehalose , the major circulating form of sugar in flies , were also rescued by either gd2 or gd2HF ( Figure 1D ) . However , the C119Y missense mutant form of gd2HF failed to rescue developmental delay , wing length or trehalose phenotypes in Ilp2–3 , 5 flies ( Figures 1C and 1D , Figure S2 ) . Thus , Ilp2HF rescues severe insulin-deficiency to an extent comparable to native Ilp2 , providing a unique example of a dual epitope-tagged insulin that retains in vivo biological activity that is nearly indistinguishable from native insulin . To investigate the physiological regulation of in vivo Ilp2 levels , we introduced a single copy of the gd2HF genomic rescue fragment by site-directed insertion into Ilp2 null mutants ( hereafter Ilp21 gd2HF ) , thereby replacing endogenous Ilp2 with Ilp2HF in the genome . Immunostaining revealed Ilp2HF protein was restricted to adult IPCs of the pars intercerebralis without detectable ectopic expression in Ilp21 gd2HF brains ( Figure 1E ) . Circulating trehalose levels were indistinguishable in Ilp21 gd2HF adults and controls ( Figure 2A ) . Quantitative reverse-transcriptase polymerase chain reaction ( qPCR ) revealed that Ilp2 mRNA levels in Ilp21 gd2HF adults and controls were significantly reduced in 3 day-old flies compared to 1 day-old flies ( Figure 2B ) . Thus , in vivo regulation of the gd2HF genomic rescue fragment recapitulates that of native Ilp2 and produces the same physiological responses . To quantify total and circulating Ilp2HF in adult flies we developed an ELISA based on commercially available monoclonal antibodies and peptide standards harboring both HA- and FLAG- epitope tags ( Figure 2C ) . This assay detected signal in sample volumes of 1 µl in a standard range of 40 pM to 4 nM ( Figure 2C; Materials and Methods ) . In contrast to mRNA levels , total Ilp2HF content of 1 and 3 day-old homozygous Ilp21 gd2HF adults did not change ( Figure 2D ) , demonstrating that changes of Ilp2 mRNA levels do not strictly correlate with changes in total protein levels . The average circulating Ilp2HF concentration in hemolymph from 1 day-old homozygous Ilp21 gd2HF adults was 100 pM and increased to 350 pM in 3 day-old adults ( Figure 2E ) , demonstrating further that levels of secreted Ilp2HF protein in hemolymph from adult flies are regulated independently of total Ilp2HF content . Based on an estimated adult hemolymph volume of 80 nanoliter [16] , we determined that the total circulating Ilp2HF rises from 0 . 05 pg in 1 day-old adult flies to 0 . 22 pg in 3 day-old adult flies . Thus , we calculate that only 0 . 1% of total Ilp2HF circulates in the hemolymph of 1 day-old flies , increasing to 0 . 35% of total Ilp2HF content in 3 day-old flies ( Figure 2E ) . Together these results indicate that a small fraction of total Ilp2HF in IPCs is secreted into the hemolymph in vivo , and demonstrate that our ELISA method permits assessment of physiological regulation of insulin production and secretion in flies . To further assess the fraction of Ilp2 secreted upon stimulation of IPCs , we isolated heads from 3 day-old homozygous Ilp21 gd2HF flies , and stimulated them with 100 mM KCl , as previously reported [7] . Under physiological conditions in 3 mM KCl adult hemolymph-like solution ( AHLS ) , 0 . 2 pg of Ilp2HF ( per head ) was secreted . This increased 7-fold to 1 . 4 pg of Ilp2HF ( per head ) upon stimulation with 100 mM KCl AHLS ( Figure 2F ) . Similar to results in vivo , only 0 . 6% of the total Ilp2 content in heads was secreted in physiological AHLS and the fractional secretion increased to 2 . 8% of the total when heads were stimulated with 100 mM KCl , comparable to the fraction of insulin secreted from stimulated rat islets [17] . These results emphasize that only a small fraction of total Ilp2 content of IPCs is secreted , even after maximal depolarization induced by 100 mM KCl . To assess if nutrient availability acutely regulates insulin secretion and circulating levels in hemolymph , as indicated by prior studies [7] , we also measured Ilp2HF in fasted and re-fed adult flies . In 3-day old flies fasted for 24 hours then re-fed for 30 minutes , circulating Ilp2HF concentration peaked then declined ( Figure 2G ) , a pattern and timing strikingly similar to that observed in fasted and re-fed humans [18] . Thus , our assays provided measures of systemic insulin levels in flies on a time scale comparable to in vivo measures of mammalian insulin dynamics . IPCs are neurons whose stimulation by glucose evokes electrical responses [19] , [20] . Modulating IPC activity is thought to affect insulin release , but changes in circulating insulin levels have not been demonstrated . Kir2 . 1 encodes an inward-rectifying potassium channel that has been used to silence electrical activity of Drosophila neurons and neuroendocrine cells [21] . We used ‘Geneswitch’ GAL4 to express Kir2 . 1 in adult IPCs [22] , which permits mifepristone-dependent conditional gene expression , and minimizes the effects of insulin perturbation during animal growth and development . In control Ilp21 gd2HF heterozygous flies , circulating Ilp2HF levels were not affected by mifepristone feeding , and were maintained near 200 pM ( Figure 3A ) , about half of the circulating Ilp2HF level in 3 day-old Ilp21 gd2HF homozygous flies , as expected . In subsequent experiments , Ilp2HF levels in Ilp21 gd2HF heterozygous flies were measured . We found that Kir2 . 1 expression induced by mifepristone feeding in adult IPCs significantly reduced hemolymph Ilp2HF concentration without affecting cell number ( Figure 3A and B ) . These results support the postulated role of ion channel activity in regulating insulin secretion [7] , and provide direct evidence that ion channel function may couple IPC activation to circulating insulin levels . Unlike larval IPCs , adult IPCs are glucose responsive [20] . In humans , GLUT1 is a major glucose transporter of pancreatic β-cells . To further assess carbohydrate sensing in adult IPCs , we used RNAi in the IPCs to knockdown expression of the type-1 glucose transporter Glut1 [23] , a gene not previously shown to regulate IPC function . In control experiments , we observed that RNAi knockdown of Ilp2 significantly reduced Ilp2 mRNA expression , total Ilp2HF content in flies , and circulating Ilp2HF levels ( Figures 3C–E ) . Knockdown of Glut1 in IPCs severely reduced circulating Ilp2HF levels , but had no detectable effect on Ilp2 mRNA levels or total Ilp2HF content in flies ( Figures 3C–E ) . Together , these data suggest that Glut1 in Drosophila IPCs is a conserved regulator of in vivo insulin secretion . To identify uncharacterized regulators of insulin expression , production , and secretion in Drosophila IPCs , we performed loss-of-function analysis of fly genes corresponding to GWAS candidate genes for T2DM [2] , [24] . Glis3 was recently shown to be required for insulin expression in mouse islet β-cells [25] . Knockdown of lmd ( orthologue of Glis3 ) in IPCs severely reduced circulating Ilp2HF levels , total Ilp2HF content , and Ilp2 mRNA expression ( Figures 3C–E ) , suggesting that lmd may regulate Ilp2 expression , similar to the role of rodent Glis3 in regulating Ins expression . BCL11A has been associated with type 2 diabetes mellitus [24] , but prior work has not linked BCL11A to insulin regulation in mammals . In contrast to lmd , knockdown of CG9650 ( orthologue of human BCL11A ) increased circulating Ilp2HF levels without affecting Ilp2 mRNA levels or total Ilp2HF content in flies ( Figures 3C–E ) , suggesting that CG9650 may regulate Ilp2HF levels post-translationally . In mammals , genetic or acquired pathological insulin resistance provokes adaptive responses in β-cells , including enhanced insulin secretion [26] , but it was not known if such facultative responses were conserved in Drosophila . Adult flies heterozygous for the InR05545 mutant allele do not have detectable growth [27] or trehalose phenotypes ( Figure 4A ) . Remarkably , we found that circulating Ilp2HF concentration was doubled in InR05545 heterozygotes ( Figure 4B ) . We observed similar phenotypes in flies heterozygous for a loss-of-function mutation in Akt1 ( Figures 4A and 4B ) , which encodes an essential conserved regulator of insulin signaling [28] . To test whether elevated hemolymph Ilp2HF levels in InR05545 heterozygous flies derived from increased production or increased secretion , we measured total Ilp2HF content . Ilp2HF content was identical in InR05545 heterozygous flies and controls ( Figure 4C ) , indicating that hyperinsulimia in InR05545 heterozygotes results from enhanced insulin secretion , not from enhanced insulin production . These results are reminiscent of adaptive phenotypes noted in mice harboring heterozygous mutations in genes encoding insulin receptor or other insulin signaling regulators [29] , [30] . The two-fold increase in circulating Ilp2HF in InR05545 heterozygotes represents only a minor fraction of the total Ilp2HF content; thus we asked whether this subtle difference could be detected using previously established methods [7] . We could not detect differences in Ilp2 accumulation in IPCs from InR05545 heterozygotes and control flies by immunofluorescence ( Figure 4D ) , suggesting that quantification of Drosophila insulin by the Ilp2HF system permits the discovery and characterization of phenotypes not detected by semi-quantitative assays of Ilp2 secretion . To identify the tissue-specific basis of the enhanced insulin secretion phenotypes in heterozygous InR mutants , we systematically knocked down InR expression using RNAi in adult IPCs , muscle , or in fat body , an organ with functions orthologous to the liver . InR knockdown in muscle using Mef2-GAL4 did not detectably alter Ilp2HF levels ( Figure 4E ) , reminiscent of normal serum insulin levels observed in muscle-specific insulin receptor knockout ( MIRKO ) mice [31] . InR knockdown in IPCs using Ilp215-1-GAL4 decreased circulating Ilp2HF levels ( Figure 4E ) , reminiscent of insulin defects found in pancreatic β cell-specific insulin receptor knockout ( BIRKO ) mice [32] . In contrast , increased circulating Ilp2HF levels were evoked by RNAi-mediated InR knockdown in the adult fat body using Lk6-GAL4 or ppl-GAL4 ( Figure 4E ) without detectable effects on total Ilp2HF content in fat body-specific InR knockdown using ppl-GAL4 ( Figure 4F ) . RNAi-mediated InR knockdown in fat body was confirmed by qPCR of fat body cDNAs ( Figure S3 ) . These results suggest that insulin secretion , not production , from IPCs is regulated by impaired insulin signaling in fat body . Thus , similar to mice with conditional insulin receptor loss in liver ( LIRKO ) [33] , targeted impairment of insulin signaling in Drosophila fat body produced enhanced insulin secretion from IPCs .
In vivo measures of circulating insulin and other peptide hormones in organisms with powerful experimental advantages , like Drosophila , could transform the scope of physiological and genetic approaches possible in these systems , and advance their use for metabolic and genomic studies . Active insulin is produced from multiple post-translational processing steps , including proteolytic cleavage and extensive disulfide bonding , and modification of a single amino acid in insulin protein can significantly alter or eliminate its hormone activity . Thus , despite intensive efforts , labeling of insulin with useful peptide epitopes while preserving in vivo hormone function has remained a challenge . We exploited quantitative synthetic lethality tests in flies to screen multiple modifications in the Ilp2 protein , and found that Ilp2 tolerated epitope insertions only at specific locations while preserving bioactivity , specifically the HA-epitope at the B-chain carboxy-terminus and the FLAG-epitope at the A-chain amino-terminus . While structural analysis for Ilp2 is not available , to our knowledge , the structure of the related insulin-like peptide Ilp5 has been reported [6] , revealing a disordered B-chain carboxy-terminus adjacent to the A-chain amino-terminus . To the extent that similar features may be found in Ilp2 , we speculate that this structural feature may be permissive for Ilp2 epitope tagging while preserving function . If so , epitope-tagging methods described here may be used to quantify and investigate function of other processed circulating peptide hormones in Drosophila , or other species . We also found that Ilp2HF bioactivity is impaired by introduction of an “Akita” missense mutation , analogous to mutations previously shown to disrupt post-translational insulin processing in rodents , and in humans with dominant mutant proinsulin syndrome [14] , [15] . This raises the likelihood that conserved mechanisms may underlie prepro-Ilp2 processing and folding in Drosophila IPCs , and that the Ilp2–3 , 5 gd2HF . C119Y line may provide a useful model for studies of protein-folding in Drosophila . Our studies revealed that Drosophila insulin expression , production , and secretion are dynamic and independently regulated in IPCs . By contrast , intracellular immunofluorescence methods that infer IPC secretion responses do not discriminate between insulin expression , production , and secretion . Moreover we also found that , upon IPC stimulation , only a small fraction of the total Ilp2 in IPCs is secreted in vivo and in vitro . Based on synthesized peptide standards , we found the circulating Ilp2HF concentration increases from 100 pM in 1 day-old flies to 350 pM in 3 day-old flies without a change of total Ilp2 content during this period . Although Ilp2 affinity for the Drosophila insulin receptor has not been reported , competitive binding studies of purified Ilp5 revealed a Kd of 350–760 pM [6] , consistent with our in vivo findings . Since distinct Ilps produced in IPCs may be co-released , Ilp2 levels may indirectly reflect release of Ilp3 and Ilp5 from IPCs . While the basis for enhanced Ilp2HF secretion in 3 day old flies is not yet known , feeding behavior may change over this period and underlie this effect . Alternatively , stimulus-secretion coupling mechanisms in IPCs may mature in the first 3 days . Both possibilities have been previously observed during the postnatal weaning and maturation period in mammals . Thus , a scalable and highly sensitive method of measuring insulin content and secretion should enable a new class of physiological studies in Drosophila , permitting genetic dissection of feeding behaviors and diet effects on insulin signaling . Our system revealed molecular and cellular mechanisms for two fly orthologues of T2DM risk genes in regulating systemic insulin levels . Glis3 was recently shown to be required for insulin expression in mouse islet β-cells [25] . Consistent with this finding , we found that IPC knockdown of lmd , a fly orthologue of human GLIS3 , reduced Ilp2 mRNA and total Ilp2HF protein levels , suggesting conserved mechanisms regulating insulin expression . In contrast , we found that CG9650 knockdown in IPCs increased circulating Ilp2HF levels , without affecting Ilp2 expression or production . Thus , the product of CG9650 likely regulates circulating insulin levels at a post-translational step . Prior studies suggested that CG9650 encodes a transcription factor with roles in axon guidance , Notch signaling and oxidative stress responses [34]–[36] , but did not identify roles in insulin processing or secretion . Likewise BCL11A , a human orthologue of CG9650 , has been associated with type 2 diabetes mellitus , but prior work has not linked BCL11A to insulin regulation in mammals . In addition , our system now permits further studies of circulating signals or neurotransmitters thought to regulate insulin secretion by IPCs , including Ilp6 , Unpaired 2 , and serotonin [9] , [37] , [38] . Thus , the ability of our system to measure insulin production and secretion permits mechanistic evaluation and linkage of candidate human diabetes susceptibility genes to roles in insulin expression , post-translational processing , or secretion . Using our system , we also detected adaptive enhancement of insulin secretion in flies with heterozygous InR or Akt1 mutations . These phenotypes are similar to those reported in mice with IR or IRS deficiency [29] , [30] , in which impaired insulin signaling in peripheral tissues promotes a ‘pre-diabetic’ condition with adaptive hyperinsulinemia compensating for systemic insulin resistance while maintaining normoglycemia . The detection and quantification of haploinsufficiency phenotypes in heterozygous InR or Akt1 mutants suggests that genetic screens using deficiency lines could identify novel regulators of insulin production and secretion . We also observed changes in circulating Ilp2 levels after specific knockdown of the insulin receptor in Drosophila fat body or IPCs , but not in muscle . These results are consistent with prior reports that fat body signals might regulate IPCs [7] , and suggest a role for fat body insulin-signaling in feedback regulation of systemic insulin levels in Drosophila . The changes in circulating Ilp2HF levels after InR knockdown in fly fat body , IPCs or muscle , were remarkably similar to changes in serum insulin observed after tissue-specific knock-out of insulin receptor in mouse liver , pancreatic β-cells or muscle [31]–[33] , the so-called LIRKO , BIRKO and MIRKO mice . Thus , integrated analyses permitted by our assays revealed that mechanisms governing facultative adaptation to pathological states like impaired insulin signaling in multiple target organs are maintained from insects to mammals . We speculate that in vivo Ilp2HF quantification in Drosophila should be useful to identify conserved regulators of insulin expression , secretion and responsiveness relevant to human health and diseases .
y1 w1118 ( Bloomington stock ID #6598 ) , Ilp21 ( #30881 ) , Df ( 3L ) Ilp2–3 , Ilp53/TM3 ( #30889 ) , InR05545/TM3 ( #1161 ) , Act5C-GAL4/CyO ( #4414 ) , Mef2-GAL4 ( #27390 ) , Lk6-GAL4 ( #8614 ) , UAS-CD4-tdTomato ( #35841 ) , UAS-Kir2 . 1-eGFP ( #6596 ) , UAS-mCherry . RNAi ( #35785; used as a control RNAi ) , UAS-InR . RNAi ( #31594 ) , UAS-Ilp2 . RNAi ( #31068 ) , UAS-lmd . RNAi ( #42871 ) , UAS-CG9650 . RNAi ( #26713 ) , and UAS-Glut1 . RNAi ( #40904 ) used in this study were obtained from Bloomington Stock Center . Drosophila orthologues for human genes were identified by Ensemble release 73 . Ilp215-1-GAL4 used in this study is made from pIlp215-1-GAL4 construct ( See below ) , and its adult expression is restricted in IPCs . Akt11/TM3 was provided by Dr . Clive Wilson ( University of Oxford ) . UAS-FLAG-dilp2 [12] was provided by Dr . Matt Scott ( Stanford University ) . ppl-GAL4 [39] was provided by Dr . Michael Pankratz ( Universität Bonn ) . Ilp2-GeneSwitch was provided by Dr . Yih-Woei C . Fridell ( University of Connecticut ) . dilp215-1-HStinger was previously described [40] . In all experiments , animals were either fed on cornmeal/dextrose/yeast food ad libitum , fasted on 1% agar only food , or re-fed on 2M glucose in 1% agar with 0 . 05% bromophenol blue for oral glucose-stimulated insulin secretion experiments at 22°C . Standard Drosophila cornmeal/dextrose/yeast food was prepared with the recipe: 1% ( w/v ) agar , 5% ( w/v ) cornmeal , 10% ( w/v ) dextrose , and 2 . 5% ( w/v ) baker's yeast . Please note that 10% ( w/v ) dextrose is about 555 mM . Mifeprestone ( Sigma-Aldrich M8046 ) was added at 0 . 2 mM when needed . pIlp215-1-GAL4 was generated by subcloning the 541 bp sequence upstream of the Ilp2 transcription start site [4] into pPTGAL . pUAST-Ilp2 was generated by subcloning 705 bp EcoR1-Xho1 fragment from DGC clone GH11579 . pUAST-Ilp2HF was generated by PCR-based site-directed mutagenesis to add 5′-TAT CCA TAT GAT GTT CCT GAC TAT GCT-3′ ( encoding the amino acids YPYDVPDYA ) sequence after the end of Ilp2 B-chain and 5′-GAT TAT AAG GAC GAC GAT GAC AAG-3′ ( encoding the amino acids DYKDDDDK ) sequence before the beginning of Ilp2 A-chain ( See Figure 1A ) . P-element mediated germline transformations were carried out to generate Ilp215-1-GAL4 , UAS-Ilp2 , and UAS-Ilp2HF transgenic lines . 2413 bp genomic Ilp2 region was amplified from y1 w1118 genomic DNA using 5′-CCGAGAATTCACACTTGGCCAACACACACACATTCATTA-3′ and 5′-ACTGTCTAGAATTGGCCAACTTGATTGGTAATGAAACGG-3′ primers and subcloned to EcoR1 and Xba1 sites on pBDP2 ( a modified version of pBDP with EcoR1 , Xba1 , and Not1 cloning sites ) [41] to generate pBDP2-gd2 . pBDP2-gd2HF was generated by replacing Ilp2 coding region in pBPD2-gd2 with Ilp2HF ORF . pBDP2-gd2HF . C119Y was generated by PCR-based site-directed mutagenesis to change from 5′-TGCTGCAA-3′ to 5′-TGTTATAA-3′ . phiC31 integrase-mediated germline transformations were carried out to generate gd2 ( attP2 ) , gd2HF ( attP2 ) , and gd2HF . C119Y ( attP2 ) transgenic lines using Bloomington stock #25710 . gd2 ( attP2 ) , gd2HF ( attP2 ) , or gd2HF . C119Y ( attP2 ) transgene was recombined to Df ( 3L ) Ilp2–3 , Ilp53 mutant backgrounds to assess phenotypic rescue of Ilp2–3 , 5 deficiency mutant . To replace endogenous Ilp2 gene with gd2HF ( attP2 ) in the genome , the gd2HF ( attP2 ) transgene was recombined into Ilp21 mutant chromosome to generate the y1 w1118; Ilp21 gd2HF ( attP2 ) strain which was used to measure the circulating Ilp2HF in hemolymph . Please note that the y1 w1118; Ilp21 gd2HF ( attP2 ) strain is homozygous for gd2HF , and their circulating llp2HF levels are 300–400 pM ( Figure 2E ) . To express Kir2 . 1 in insulin producing cells conditionally , the Ilp2-GeneSwitch/CyO; Ilp21 gd2HF ( attP2 ) dilp215-1-HStinger strain was crossed to flies harboring transgene encoding UAS-Kir2 . 1-eGFP , and appropriate progeny were fed 200 µM Mifeprestone or vehicle ( ethanol ) in cornmeal/dextrose/yeast food for 48 hours . The progeny carry only one copy of gd2HF , and their circulating Ilp2HF levels are 100–200 pM ( Figure 3A ) . To knockdown genes in adult IPCs and measure Ilp2HF in hemolymph , TRiP RNAi lines were crossed to the UAS-Dcr-2 . D; Ilp21 gd2HF ( attP2 ) Ilp215-1-GAL4 strain . To knockdown genes in adult muscles and measure circulating Ilp2HF in hemolymph , TRiP RNAi lines were crossed to the UAS-Dcr-2 . D; Ilp21 gd2HF ( attP2 ) Mef2-GAL4 strain . To knockdown genes in adult fat body tissues and measure circulating Ilp2HF in hemolymph , TRiP RNAi lines were crossed to either the UAS-Dcr-2 . D; Ilp21 gd2HF ( attP2 ) Lk6-GAL4 or the ppl-GAL4 UAS-Dcr-2 . D; Ilp21 gd2HF ( attP2 ) strain . Progeny from these crosses carry only one copy of gd2HF , and their circulating Ilp2HF level are 100–200 pM ( Figure 3E and Figure 4B and E ) . One wing per female fly was dissected in isopropanol , mounted in Canada balsam:Methyl salicylate ( 4∶1 ) on a slide , and heated on 65°C hot plate for 1 hour to harden the mounting media . The distance between the distal end of the L3 wing vein and the posterior end of wing hinge was measured using AxioVision software . 5 wing spans were measured per genotype , and statistical differences between genotypes were determined with a two-tailed Student's t-test . The results are presented as the mean ± standard deviation . Immunostaining of adult brains was performed as described [40] with modifications: Affinity purified rabbit polyclonal anti-Dilp2 antibody ( 0 . 5 µg/ml ) , mouse monoclonal anti-FLAG M2 antibody ( 1 µg/ml; Sigma-Aldrich F1804 ) , Rat monoclonal anti-HA 3F10 antibody ( 0 . 1 µg/ml; Roche 1867423 ) , and Alexa Fluor 488 , 547 , and 647 secondary antibodies ( 2 µg/ml; Life Technologies ) were diluted and incubated in PBS with 0 . 3% Triton-X100 . Confocal laser scanning microscope images were obtained using Leica TCS SP5 or SP8 . Accumulation Ilp2 in IPCs was quantified as described previously [7] . Adult brains were dissected and stained for immunofluorescence as described above , and mounted with the IPCs oriented towards the coverslip . Confocal imaging parameters were optimized such that images of all samples could be acquired within the dynamic range of constant laser and scan settings . Confocal Z stacks of the IPCs were acquired with a step size of 1 µm . For quantification , image stacks were summed and the mean-pixel intensity of a region of interest ( ROI ) containing the entire IPC cluster was measured and subjected to background subtraction using an ROI drawn adjacent to the cell cluster . Average mean pixel intensity of IPCs across biological replicate brains for each condition is expressed in arbitrary units ( a . u . ) . In all assays , three separate samples per specific fly group or condition were collected . Unless otherwise noted , 3-day-old male flies fed ad libitum were used in all experiments . We avoided using female flies due to possible feeding behavior changes in virgin and mated females [42] , and larger observed standard deviations in all metabolic assays we used . All flies were transferred to vials with fresh food 24 hours prior to hemolymph collection to ensure similar nutritional conditions except fasted groups which were maintained on vials with 1% agar for 24 hours prior to the collection . In acute re-feeding and re-fasting experiments , fasted flies were placed on 2M glucose in 1% agar with 0 . 05% bromophenol blue for 30 minutes , then re-fasted flies . Because not all 24-hour fasted flies commence feeding on 2M glucose , only flies with visibly ingested blue food coloring in their gut were selected for hemolymph sampling . To elute hemolymph , sixty male flies per group were placed in a modified Zymo-Spin IIIC column ( Zymo Research Corporation C1006 ) in which DNA-binding filter were removed and thoroughly washed with water . The column containing male flies was centrifuged twice at 9 , 000 g for 5 minutes at 4°C . This yielded approximately 1 . 5 µl hemolymph , which was used for either trehalose assays or ELISA . A single fly was placed in a 1 . 5 mL centrifuge tube with 100 µl of PBS containing 1% Triton X-100 . Four samples were prepared for each genotype . The samples were ground using a pestle and cordless motor ( VWR 47747-370 ) , and lysed at room temperature for 30 minutes on a rotary shaker . The lysed samples were centrifuged at 21 , 000 g for 5 minutes at room temperature and 50 µl supernatant from the centrifuged samples were used for ELISA . Adult hemolyph-like saline ( AHLS ) was prepared with the recipe: 2 mM CaCl2 , 3 mM KCl , 8 . 2 mM MgCl2 , 108 mM NaCl , 4 mM NaHCO3 , 1 mM NaH2PO4 , 3 mM Glucose , and 2% bovine serum albumin . Heads of 3 day-old Ilp21 gd2HF males were carefully separated from bodies in AHLS to maintain foregut and crop . 15 isolated heads were transferred to a centrifuge tube containing 100 µl of AHLS , and allowed to recover for 1 hour at room temperature . Three samples per condition were prepared . The samples were washed 3 times with AHLS , and incubated in 100 µl of AHLS containing either 3 mM or 100 mM KCl for 30 minutes . 100 µl of the incubated AHLS from the samples was saved and 50 µl was used for ELISA . To measure total content , 150 µl of PBS containing 1% Triton X-100 was added to the remaining heads in the tubes , and the samples were ground using a pestle and cordless motor . After 30 minutes lysis at room temperature , the samples were centrifuged at 21 , 000 g for 5 minutes . 10 µl supernatant from the centrifuged samples was diluted in 90 µl PBS , and 50 µl of the diluted sample were used for ELISA . We coated wells in Nunc-Immuno modules ( Thermo Scientific 468667 ) with 100 µl of anti-FLAG antibody ( Sigma-Aldrich F1804 ) diluted in 0 . 2 M sodium carbonate/bicarbonate buffer ( pH9 . 4 ) to a final concentration of 2 . 5 µg/ml , then incubated for 16 hours at 4°C . The plate was washed twice with PBS containing 0 . 2% Tween 20 ( PBTw0 . 2 ) , then coated with 350 µl of PBS containing 2% bovine serum albumin ( Fisher Scientific BP1600 ) for 16 hours at 4°C . The plate was washed three times with PBTw0 . 2 . For circulating Ilp2HF measurement , 1 µl of hemolymph or 1 µl of FLAG ( GS ) HA peptide standards ( DYKDDDDKGGGGSYPYDVPDYAamide , 2412 daltons: LifeTein LLC ) at 0 . 1–10 pg/µl was mixed with 50 µl of PBS containing 1% Triton X-100 and 5 ng/ml anti-HA-Peroxidase 3F10 antibody ( Roche 12013819001 ) , vortexed , centrifuged briefly , and transferred to wells on the plate . For total Ilp2HF content measurement , 50 µl of supernatant from single fly lysate or 50 µl of FLAG ( GS ) HA peptide standards at 5–500 fg/µl was mixed with 50 µl of PBS containing 1% Triton X-100 and 5 ng/ml anti-HA-Peroxidase 3F10 antibody , vortexed , centrifuged briefly , and transferred to wells on the plate . For samples derived from in vitro head assays , 50 µl of collected media or diluted head lysate was used . The wells were sealed with an adhesive film ( Thermo Scientific 232698 ) , and incubated in a humidity chamber for 16 hours at 4°C . We removed the samples by aspirating , and washed the wells six times with PBTw0 . 2 . 100 µl of 1-Step Ultra TMB ELISA Substrate ( Thermo Scientific 34029 ) was added to each well and incubated on a rotary shaker for 30 minutes at room temperature . The reaction was stopped by adding 100 µl of 2 M sulfuric acid , and the absorbance at 450 nm ( A450 ) was immediately measured on a SpectraMax M5 ( Molecular Devices ) . To convert concentration to mass in a given volume , we used a molecular weight of 7829 daltons for mature Ilp2HF protein . 1 µl eluted sample from the centrifuged flies was diluted in 9 µl PBS , vortexed , centrifuged briefly , and heated at 70°C for 5 minutes to inactivate endogenous Trehalase . 2 µl of the heated sample was added to 200 µl of Glucose Hexokinase Reagent ( Thermo Scientific TR15421 ) with or without Porcine Kidney Trehalase ( 1∶1000; Sigma-Aldrich T8778-5UN ) , incubated at 37°C for 16 hours , and the absorbance at 340 nm ( A340 ) was measured on a SpectraMax M5 . The trehalose concentration in the sample was determined by subtracting the glucose concentration from the total sugar concentration . Four female flies per group with three biological replicates were homogenized in 600 µl of TRIzol Reagent ( Life Technologies 15596-018 ) , and total RNA was isolated according to the manufacturer's protocol . To isolate total RNA from larval fat body , we dissected fat body tissues from 6 larva per group . Three biological replicates were homogenized in 600 µl of TRIzol Reagent . Total RNA pellet was resuspended in 30 µl of water . 1 µg of total RNA was treated with DNAse I , heat-inactivated , and reverse transcribed in 10 µl reaction using High Capacity cDNA Reverse Transcription Kit ( Applied Biosystems 4368814 ) . 1 . 5 µl of cDNA was used in a final volume of 15 µl for quantitative PCR reaction ( Solaris qPCR Low ROX Master Mix , Thermo Scientific AB-4352/C ) , and PCR amplification was detected by 7500 Real Time PCR system ( Applied Biosystems ) . Relative expression levels of Ilp2 or Ilp2HF were determined by Applied Biosystems Taqman probe for Ilp2 ( Dm01822534_g1 ) . Relative expression levels of InR were determined by Applied Biosystems Taqman probe for InR ( Dm02136224_g1 ) . Applied Biosystems Taqman probe for Rpl32 ( Dm02151827_g1 ) was used as the internal control to determine relative expression of Ilp2 and InR . | Genome-wide association studies in patients with type 2 diabetes mellitus have identified more than 65 loci , encoding up to 500 candidate susceptibility genes . Thus , investigators are fundamentally challenged to ( i ) screen and identify relevant candidates in vivo , ( ii ) determine if loss- or gain-of-function underlies the association , ( iii ) link perturbed gene function to hallmark type 2 diabetes mellitus physiological phenotypes like insulin production or secretion , and ( iv ) identify relevant tissue ( s ) where the biological function of a specific regulator is required . Here we exploit Drosophila genetics to reveal the molecular functions of evolutionally conserved regulators that are associated with human type 2 diabetes mellitus . Targeted knockdown of Drosophila orthologues of diabetes risk genes revealed tissue-specific roles for these genes in regulating insulin production and secretion . These findings should accelerate use of Drosophila and other genetically-tractable systems to discover conserved mechanisms and regulators controlling in vivo insulin dynamics relevant to diabetes and other human diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"of",
"disease",
"genetics",
"biology",
"and",
"life",
"sciences"
] | 2014 | A Genetic Strategy to Measure Circulating Drosophila Insulin Reveals Genes Regulating Insulin Production and Secretion |
The modular organization of networks of individual neurons interwoven through synapses has not been fully explored due to the incredible complexity of the connectivity architecture . Here we use the modularity-based community detection method for directed , weighted networks to examine hierarchically organized modules in the complete wiring diagram ( connectome ) of Caenorhabditis elegans ( C . elegans ) and to investigate their topological properties . Incorporating bilateral symmetry of the network as an important cue for proper cluster assignment , we identified anatomical clusters in the C . elegans connectome , including a body-spanning cluster , which correspond to experimentally identified functional circuits . Moreover , the hierarchical organization of the five clusters explains the systemic cooperation ( e . g . , mechanosensation , chemosensation , and navigation ) that occurs among the structurally segregated biological circuits to produce higher-order complex behaviors .
The brain consists of a remarkably complex hierarchical structure ranging from ion channels of individual neurons to systemic neuronal networks of subsystems responsible for specific functions . To perform natural computation efficiently , the brain has evolved to have specialized modules with locally dense connections to integrate functions and produce complex behaviors . Because brain structure is closely related to function , an understanding of the topological structure of neuronal organization in the brain is crucial for insight into how neuronal networks perform their precise functions [1] , [2] , [3] , . To uncover the neurobiological mechanisms of brain functions , mapping of the complete wiring diagram of a neural system has been attempted; this field is called connectomics [2] , [9] . Although connectomics is presently at an early stage and data mining related to its application has only recently begun , the connectomics approach may eventually shed light on the fundamental principles underlying brain functions and the pathological mechanisms of neuropsychiatric disorders that arise from faulty wiring , such as schizophrenia and autism [2] , [5] , [6] , [9] , [10] , [11] , [12] . As accurate large-scale data describing the topology of networks become available in various fields , complex network analysis tools have been developed and applied . The study of complex networks involves the investigation of important topological features of a network with connections among its nodes that are neither purely regular nor purely random . This technique has been applied to complex networks of the real world , such as the worldwide web [13] , metabolic networks [13] , food webs [13] , and neural [2] , [5] , [7] and social networks [2] , [5] , [7] , [13] , [14] . These complex networks have shown universal structural features including small-world properties [13] , [14] , power-law degree distributions [13] , the existence of repeated local motifs [2] , [15] , and robustness and fragility against attacks [13] . Recently , the brain , a typical example of a complex network , was found to exhibit small-world topology from the microscopic level ( e . g . , the neuronal network of C . elegans ) [14] , [16] to the macroscopic level [2] , [16] , [17] , [18] . Scale-free degree distributions are observed in fMRI-based voxel networks of human brains [2] , and structural and functional motifs can be detected in the large-scale cortical networks of macaque monkeys and cats [2] . Robustness and fragility of brain structural networks with respect to lesions and diseases have also been examined quantitatively [7] , [12] , [18] , [19] . Another significant issue in complex network analysis is the determination and characterization of the hierarchical cluster structure in a network , i . e . , the appearance of densely connected groups of nodes with sparser connections among groups and their association at higher levels [20] , [21] , [22] . Topological clusters in brain structure may correspond to sets of distinct anatomical modules of neurons [2] , [5] , [6] , [7] , [23] , [24] , [25] . Detection of cluster structure in the brain is of critical importance because it provides valuable clues regarding the relationship between anatomical clusters and functional circuits . Such a relationship is based on the modular view of network dynamics , which assumes that different groups of neurons perform different functions with some degree of independence . Several studies have investigated the large-scale network structure of the mammalian cortex and its association with cortical function . Both the structure as a whole [2] , [6] , [7] , [23] , [25] and subsystems [24] of the brain have several distinct anatomical substrates ( segregation ) as well as functional connectivity ( integration ) , implying an intimate association between structural clusters and functional modules at the macroscopic level [8] , [17] , [18] . However , because of the complexity of the connectivity architecture at the level of individual neurons , no studies have reported whether the connectome of an entire nervous system exhibits a hierarchical cluster structure . Therefore , the aim of this study was to investigate the possible existence of cluster structure in the neuronal network of the entire nervous system of the nematode Caenorhabditis elegans ( C . elegans ) using the updated version of its wiring diagram ( connectome ) based on synaptic connection topology . The microscopic worm C . elegans has 302 neurons with approximately 8 , 000 synapses and is the only model organism in which the wiring diagram of the entire nervous system is almost completely known [3] , [26] . We utilized this connectome to determine whether a network of individual neurons exhibits hierarchical cluster structure with non-uniform synaptic connections or a random network structure with homogeneous synaptic connections . To detect a possible hierarchical cluster structure in the C . elegans connectome , we used the modularity-based community detection algorithm for directed weighted networks [20] , [27] . Modularity is a quantitative measure defined as the number of edges falling within groups minus the expected number in an equivalent network with edges placed at random; positive values demonstrate the possible presence of cluster structure [20] , [22] , [27] . A significant advantage of the modularity-based community detection algorithm is that it can show a network to be indivisible ( i . e . , that it contains no cluster structure ) if no true division of the network results in a positive modularity . Because a biological neural network is inherently directed and weighted , we implemented a recently introduced version of modularity function for directed and weighted networks and applied it to the directed weighted C . elegans connectome [27] . Although the modularity maximization approach of community detection has become the most popular and powerful method in the discipline , several recent studies have addressed some problems with this method [28] , [29] . Because modularity optimization is known as an NP-complete problem , researchers have used a set of approximation heuristics to obtain a near-optimal community assignment vector without knowing the overall properties of the modularity landscape . However , Good et al . [29] examined the presence of an extremely rugged structure around the top of the modularity landscape through extensive computational validation of modular properties in many popular networks . This finding implies that the modularity maximization method may provide a great number of near-optimal vectors with very inhomogeneous characteristics and may not permit the determination of the goodness of each community vector without prior non-topological knowledge about node characteristics [28] , [29] . In the case of the C . elegans connectome , however , we have a valid cue to overcome this issue: the information given by the bilateral functional symmetry of the neuronal cells as a constraint for optimization . Thus , we first show that the conventional implementation of modularity maximization using the spectral method and another popular greedy algorithm cannot produce biologically valid community assignment vectors . Second , we propose a novel scheme for constrained modularity optimization using a simulated annealing procedure . As a stochastic optimization method , this procedure allows a comparison of a diverse set of community assignment vectors for identification of a near-optimal partition . Through the extensive computational task of producing various community assignment vectors , we finally achieved a stable vector with the highest modularity value under given biological constraints . After detecting topological clusters in the C . elegans connectome , we investigated their network properties including spatial distribution of the neurons within clusters and their association with experimentally identified functional circuits .
We analyzed the one-dimensional spatial representation of the C . elegans wiring diagram recently published by Chen et al . [30] and Varshney et al . [31] , which was updated from the dataset of White et al . [3] where connections were identified by electron microscopic reconstructions . The data contained information on the direction and number of connections via chemical synapses and electrical junctions among neurons in the entire nervous system as well as one-dimensional spatial positions of neurons ( i . e . , somal centers ) along the anterior-posterior body axis . All connections between non-pharyngeal neurons were included except those of CANL/R and VC6 , which did not have obvious synapses . Consequently , the model connectome had 279 neurons ( pharyngeal and unconnected neurons excluded ) with 6 , 393 chemical synapses and 890 electrical junctions . Data sets are available at http://www . wormatlas . org/neuronalwiring . html . In this study , the complete neuronal wiring diagram of C . elegans through chemical synapses and electrical junctions ( connectome ) was considered as a directed weighted network with basic topological attributes including degree , weight , and strength [32] . The degree equals the number of synaptic partner neurons of a neuron and the weight is the appropriate sum of synapses between specific neuronal partners . The strength represents the total weights of synaptic connections afferent to or efferent from a neuron . A weighted asymmetric adjacency matrix was devised to illustrate the synaptic connections between 279 neurons . The matrix size was accordingly 279×279 and the sum of the weights of each element represented the number of synapses from one neuron to another . The summed weight of all elements in the adjacency matrix ( the total number of chemical synapses + double the total number of electrical junctions ) was 8171 . To identify possible cluster structures in the C . elegans connectome , we used the modularity-based community detection algorithm for a directed and weighted network . The modularity value , Q , indicates the degree to which a given partition succeeds in maximizing intra-cluster weights and minimizing inter-cluster weights compared to a null model given a strength sequence . To detect clusters in a directed and weighted network , we implemented a directed network version of modularity , which is defined as follows: ( 1 ) ( 2 ) where A is the adjacency matrix of a directed weighted network , Siin and Siout indicate incoming and outgoing strengths , respectively , of neuron i and is the global sum of the weights of all dyads . Hence , Bij becomes a measure of the extent to which the number of connections from neuron j to neuron i are prominent in comparison with a randomized network . After achieving the modularity function , we needed to search for a community assignment vector s that approximates the global maximal value of Q . To prevent the generation of suboptimal outcomes when using several deterministic algorithms , we implemented a stochastic hill climbing approach [21] to validate diverse near-optimal values . The algorithm is designed following the standard scheme of a metropolis algorithm , setting the objective function as a modularity function . First , we randomly assigned groups of nodes and flipped each nodal membership depending on computational temperature , T , and the marginal modularity gained by this action . As this optimization procedure repeats , T decreases so that we can search more limited areas with higher modularity values . After achieving an optimized vector with this individual nodal level manipulation , we repeated the same metropolis procedure in the level of communities . That is , merging two clusters with respect to the modularity gain . This method is the most accurate to date and contains assignment vectors in its pool of solutions that can be achieved by other community detection methods [21] , [28] . In addition to this standard procedure , we considered an optimization method with external constraints . Using the information given by non-topological prior knowledge , we constrained the type of solutions [28] , [29] . In the present study , the given constraint is the bilateral symmetry of neurons , which indicates that each bilateral pair should be classified in the same cluster . Thus , we searched for community assignment vectors within the global modularity landscape that satisfied this condition . This additional term can be easily implemented in the algorithm by providing a simultaneous cluster membership change constraint for each bilateral pair . We performed an additional analysis on the proximity of the obtained clusters . Following the second phase optimization procedure introduced in the fast unfolding algorithm [29] , we built a new network whose nodes consisted of communities found by the initial simulated annealing algorithm and the link weights between the newly assigned nodes ( i . e . , summed values between inter-cluster weights ) . We then applied the same modularity maximization approach as described previously . This procedure revealed clusters of clusters where significant levels of clustering were present between previously obtained clusters .
To examine the presence of hierarchical cluster structure in the C . elegans connectome , we first estimated the modularity of this connectome within a framework of modularity-based community detection [20] , [27] . This method seeks optimal divisions of the network into densely connected subgroups by maximizing the modularity Q . Because the C . elegans connectome had a power-law distribution of synaptic weights ( Figure S1 ) and synaptic directions between neuronal connections , it was necessary to include the directionality and the number of synapses among neurons in the asymmetrically weighted elements of the adjacency matrix . Although the modularity maximization approach of community detection has become a standard methodological means to detect possible community structures of networks , recent theoretical works have shown extreme degeneracy of solutions that produce near-optimal modularity values . One way to overcome this problem is to reduce the number of community assignment vectors using information given by prior knowledge of the node properties . Given that most bilateral neuronal pairs of C . elegans have similar functional roles [3] , [26] , [33] and accepting the principle of structure-function association in evolutionary biology [8] , [29] , structural clusters driven by an appropriate community detection method should not assign each member of a bilateral neuronal pair to a different structural cluster . We thus proposed a novel scheme to obtain an optimal community assignment vector . The simulated annealing method with external constraints in this study was utilized to find an optimal community assignment vector among the pool of solutions satisfying the bilateral symmetry condition . Figure 1A depicts the properties of diverse sets of solutions derived using the simulated annealing method without external constraints , the spectral detection method , and the fast unfolding algorithm . The spectral detection algorithm is one of the most popular algorithms because of its short computational time [22] . The fast unfolding algorithm is one of the most accurate and fast deterministic algorithms , resulting in a high modularity value [12] , [28] . However , Figure 1A demonstrates that the set of solutions driven by the simulated annealing method [21] produced a higher modularity value than the other two solutions . Moreover , the solutions of the two methods had low biological plausibility . The number of separated left/right pairs of each community assignment vector was 8 and 9 of the 93 bilateral neuronal pairs ( totally 186 neurons ) , respectively , whereas the simulated annealing algorithm produced solutions with less-separated bilateral pairs and comparable modularity values . Figure 1B and Figure 1C present the modularity values and the similarity of solutions derived using the simulated annealing method with external constraints . To show the stability of solutions in the high modularity region of the community assignment vectors with no separated bilateral pairs , we implemented a parameter called ‘variation of information’ that quantified the difference between two community assignment vectors . Variation of information between two partitions C and C' is defined as follows: ( 3 ) where X and Y denote the vectors representing the cluster assignment of community divisions C and C' , respectively , H ( X|Y ) is the conditional entropy indicating the amount of additional information needed to describe C given C' , and H ( Y|X ) indicates the opposite condition . Consequently , V ( C , C’ ) = 0 indicates that two partitions are exactly identical and thus do not require any additional information to describe each other whereas a higher value indicates a greater difference in community assignment [34] . Because the maximum possible value of the difference between two partitions of a network having 279 nodes in terms of V is log 279 , we rescaled the values to range from 0 to 1 by dividing the original value by log 279 [34] . Figure 1C shows that the solutions obtained using the external constraint condition exhibited stable properties in the highest modularity region ( Q>0 . 480 ) where each partition pair exhibited very low V values ( 0 . 122±0 . 002 ) . Through an extensive computational analysis ( over 10 , 000 trials of simulated annealing with external constraints ) , we obtained an optimal cluster assignment with Q = 0 . 490 , resulting in no separated bilateral neuronal pairs . This value was substantially higher than the average Q ( 0 . 283±0 . 009 ) of null networks obtained by swapping synaptic connections between neuronal pairs of the original network while preserving the out-strengths of the neurons [35] . With this maximal Q value , we found 5 distinct anatomical clusters in the C . elegans connectome . This result indicates that , among the possible connection distributions in the original strength sequence , the neuronal architecture of C . elegans exhibits a statistically significant modular structure . We also measured the topological proximity between the obtained clusters to determine whether a hierarchical relationship was present between them . Following the second phase optimization procedure of the fast unfolding algorithm , we built a new network whose nodes are communities found by the initial simulated annealing algorithm . ‘Link weights’ between the newly assigned nodes consist of summed values between inter-cluster weights . By applying the modularity maximization algorithm to this new network , we showed that the previously obtained 5 clusters further clustered into 2 clusters in the higher level . This procedure allowed us to obtain a hierarchical dendrogram of the 5 modular clusters . Former branching was assigned a nomenclature of 1 ( 2 in the left digit ) , and later branching was called 1 ( or 2 rightward ) . For instance , cluster 11 , 12 and 13 have the same mother . Out of 279 neurons , 57 neurons were in cluster 11 , 79 in cluster 12 , 14 in cluster 13 , 74 in cluster 21 , and 55 in cluster 22 . Cluster information for each neuron is listed in the Table S1 . The topological relationships based on synaptic connections within and among the clusters are demonstrated in the reordered adjacency matrix of the C . elegans connectome in Figure 2A . Although the off-diagonal elements of the adjacency matrix for inter-cluster links had low values , large values of the diagonal elements in Figure 2A indicate that most of the links were intra-cluster for each of 5 clusters . Figure 2A also provides information on the hierarchical relationship between the clusters . As illustrated , we observed many ties across the clusters that depended on hierarchical proximity: cluster 11 , 12 , and 13 formed a grand cluster and cluster 21 and 22 formed another grand cluster . The complete hierarchical dendrogram of the entire neurons , which accords with this cluster level hierarchical relationship , is presented in the supplementary information ( Figure S6 ) . The fact that the length of C . elegans is about ten times greater than its diameter allowed us to consider the positional distributions of neurons within each cluster in one dimension [3] , [26] , [30] , [33] . Figure 2B shows the average distances between the somata of neurons within each cluster and between clusters . Between inter-cluster neurons , the average distance was smaller than 0 . 5 unit length ( Figure 2B ) , whereas the two largest proximal ganglia groups ( groups of neurons aggregated based on the positions of their cell bodies ) , G1 to G3 and G6 to G10 , were located at large average distances from each other ( Figure 2C ) . While C . elegans neurons are spatially concentrated in a manner related to their ganglionic affiliation , we failed to observe a strong spatial localization of neurons belonging to the same cluster , except for those in clusters 11 and 12 . We estimated the density of the somata of all neurons on the horizontal plane along the anterior-posterior body axis of the animal ( Figure 2D ) . We found that clusters 11 and 12 were densely localized in the head . In contrast to the extreme spatial localization of ganglia ( Figure 2D ) [36] , we detected a body-spanning cluster , cluster 22 , that was distributed from the head to the tail of the worm's body ( Figure 2D ) . We also noted the presence of clusters 13 and 21 , which loosely spanned the anterior and posterior parts of the body , respectively . We examined the compositions of neuronal types and ganglionic affiliations of neurons within clusters as shown in Figure 3 . The diversity of neuronal types for a cluster was quantitatively measured using the index of qualitative variation ( IQV ) ( see SI for detailed information ) [37] . The IQV measures the heterogeneity of composition in a cluster; high IQV scores for a cluster indicate that the cluster is composed of various neuronal types or ganglionic neurons . In other words , if a set is composed of only a few dominant types , the IQV approaches 0 , and it reaches 1 in the opposite case . Except for cluster 22 , the clusters exhibited IQV values ranging from 0 . 78 to 0 . 98 , indicating that the majority of the clusters did not possess dominant neuronal types ( Figure 3A ) . In addition , four of 5 clusters did not display dominant neurotransmitter types ( Figure S2 ) . The single exception was cluster 22 , which consisted of 90% motor neurons and had an IQV value of 0 . 25 ( Figure 3B ) ( also see Figure S4 ) . All ganglia exhibited a rich diversity of cluster affiliations in their membership ( Figure 3A and C ) , indicating that low levels of overlaps exist between ganglia and cluster assignments . Quantitatively , the IQV between ganglia and cluster assignments was 0 . 36 , indicating a low level of correlation between the two assignments . Classification of nodes using their intra- and inter-cluster connections has been used for the cartographic representation of complex networks [21] . To determine whether the characteristics of neurons in the context of a modular network are associated with their biological functions , we estimated the within-module weight ( Z ) and participation coefficient ( P ) of all neurons in the C . elegans connectome . The within-module weight ( Z ) evaluates how strongly a neuron is connected to other neurons within its cluster , and the participation coefficient ( P ) quantifies how extensively the connections of a neuron are distributed among different clusters . By plotting the P and Z values for each neuron in a two-dimensional plane , we characterized each neuron as either a provincial or peripheral node , a hub , or a node with few within-module degrees ( see SI for detailed information ) . The P and Z values for each neuron are listed in the Table S3 . According to the classification criteria suggested by Guimera and Amaral [21] , we found that most of the neurons belonged to groups of ultra-peripheral nodes ( role R1 , 42 out of 279 ) , peripheral nodes ( role R2 , 196 out of 279 ) or non-hub connector nodes ( role R3 , 34 out of 279 ) ( Figure 4A ) . Neurons with the highest P values ( P>0 . 62 ) were concentrated in the non-hub connector class ( role R3 ) of low Z values ( -2<Z<2 ) rather than in the connector hub class ( Role R6 ) . This result indicates that the clusters in the C . elegans connectome are connected via internal peripheral members . Interestingly , most neurons ( 86% ) classified as ultra-peripheral nodes ( role R1 ) with P = 0 were sensory or motor neurons , whereas all of the neurons classified as connector hubs ( role R6 ) were command interneurons ( AVA , AVB , PVC ) [3] . These results suggest that interneurons play an important role both in connecting other neurons to form a cluster and in bridging between clusters . To determine whether our topological clusters have functional relevance , we investigated how topological clusters were associated with functional neural circuits already studied experimentally . In Figure 4B , we present a diagram focusing on the two circuits having the largest memberships: mechanosensation and chemosensation [26] , [38] , [39] . C . elegans responds to various mechanical cues by means of specific sensory neurons . ALM , AVM , PLM , and PVD have roles in sensing mechanical touch [40] , [41] , [42] . These mechanosensory neurons belonged to cluster 21 ( Figure 4B ) . Cluster 21 also contained some command interneurons , AVD and PVC , which are responsible for transmitting mechanosensory inputs to motor neurons [40] , [41] , [42] ( Figure 4B ) . In the case of chemosensation , chemical signals are sensed by different sets of neurons . For example , the neurons AWC and ASE have roles in sensing volatile and water-soluble compounds , respectively [43] , [44] . AIA , AIY , AIZ , and AIB are the 1st layer interneurons that receive synaptic inputs directly from sensory neurons; together with the chemosensory neurons , they belong to cluster 11 . The 1st layer interneurons direct their outputs onto the 2nd layer interneurons ( RIA , RIB , RIM , and SMB ) , which belong to clusters 11 and 12 . When chemical/mechanical signals are processed and transmitted within the C . elegans neural networks , the ultimate outcome is movement and behavior mediated by the motor neurons connected to body muscles . For instance , in chemosensation , signals processed in the 2nd layer interneurons and mechanosensory neurons pass onto motor neurons via command interneurons ( AVD and PVC ) [3] , [38] , [39] . When body muscles contract , class A motor neurons are important for backward movement , while class B motor neurons have a role in forward movement [40] , [41] , [42] ( also see Figure S5 and Table S4 ) . All of the class A and B motor neurons belonged to cluster 22 ( 13 of 21 class A and 12 of 18 class B neurons ) and cluster 21 ( 8 of 21 class A and 6 of 18 class B neurons ) . Interestingly , AVA neurons , the command interneurons that are important for backward movement [40] , [41] , [42] , and AVB neurons , [40] , [41] , [42] responsible for forward movement , belonged to cluster 21 together with some class A and B motor neurons , indicating that the body-spanning clusters ( 21 and 22 ) are responsible for forward and backward movement . Taken together , these observations suggest that the topological clusters we observed are closely associated with functional circuits in the C . elegans connectome ( Figure 4 ) . To quantitatively demonstrate the discriminative power of the current community assignment , we used a boot-strap sample t-test . The aim of this analysis was to determine whether a randomly assigned community vector with the same cluster size distribution would show a similar level of discriminative power for the circuits represented in Figure 4B as the optimized solution . By assigning the functional groups of neurons as chemosensory neurons , 1st layer interneurons , 2nd layer interneurons , mechanosensory neurons , command interneurons , and class A and B motor neurons , we measured the extent to which the original community assignment vector was consistent with the functional grouping of the 84 neurons . The resulting V value between the optimized assignment vector and the functional grouping was 0 . 348 , whereas the mean value between randomized vectors with the same cluster size distribution and the functional grouping was 0 . 893 ( ±0 . 002 ) . This result implies that the optimized vector's concordance with the functional groups was significant at the 99% confidence level . To examine whether the deduced information flow was reflected in the clusters at the level of synapse directionality , we estimated the inward/outward synapse ratio of each cluster toward other clusters . We considered that cluster 11 , the major members of which are sensory neurons , was the information-producing cluster and thus should have mostly outward synapses . Indeed , 68% of cluster 11 neurons had outward synaptic weights ( Figure 5A ) . On the contrary , cluster 22 , which was the information-receiving cluster ( i . e . , composed of motor neurons ) , had mainly inward synapses ( 65% having inward synaptic weights ) . Clusters 12 , 13 and 21 , which possessed comparable numbers of neuronal types ( clusters 12 and 21 ) or were predominantly composed of interneurons , exhibited balanced levels of inward and outward synaptic weights . To investigate the information flow between clusters in terms of complex networks , we estimated ‘hub and authority scores’ of the clusters in the C . elegans connectome . Hub and authority scores measure the quality of the connections each node contains and show the significance of nodes in a directed network in a dynamic regime ( see SI for detailed information ) [45] . In our cluster-to-cluster network analysis , a cluster with a high hub score is linked through outward synapses to clusters having many inward synapses . Conversely , authoritative clusters have many inward synapses from clusters that bridge to them through outward synapses . We found that the authority scores of the clusters were proportional to the intensity of inward synaptic weights ( Figure 5A and B ) . Thus , the body-spanning cluster 22 , whose members are predominantly motor neurons , acted as an ‘authority’ receiving information from hub clusters to produce consequential behaviors . In contrast , the hub scores of the clusters were not strongly related to their outward synaptic weights . The cluster with the highest outward weight ratio ( cluster 11 ) was not the most prestigious hub cluster , whereas the cluster with the highest hub score ( cluster 21 ) had equivalent degrees of in and out synapses . This result may reflect the presence of indirect connections from the information-producing cluster 11 to the information-receiving parts of various clusters . In contrast , clusters 21 and 22 , which exchange connections with each other , are motor neuronal clusters with direct synaptic connections ( Figure 5C ) . With respect to functional relevance , our topological analysis provided a hierarchical model of the information flow among structural clusters ( Figure 5C ) . For example , the hierarchically close clusters 11 and 12 were functionally associated with each other for chemosensory behavior ( navigation for food searching ) [39] . As noted , cluster 11 contained mostly chemosensory neurons and 1st layer interneurons , while cluster 12 contained 2nd layer interneurons and motor neurons responsible for head and neck movement . This hierarchical relevance was also apparent between clusters 21 and 22 for the behavior of anterior touch response . Interestingly , our prediction of the information processing procedure between the clusters agreed with the nodal-level depiction of information processing hierarchy derived using a Laplacian matrix analysis [31] . We computed the cluster-level mean value of information processing hierarchy introduced by Varshney et al . [31] . This measure describes a chain of information producers and receivers in a one-dimensional axis using information obtained from complex recursive structural interactions between the neurons in the connectome . As a result , information producers have a high level of parametric value and receivers have a low level of parametric value . The average values of this parameter for neurons belonging to each cluster were as follows: cluster 11 ( 0 . 65±0 . 55 ) > cluster 13 ( 0 . 17±0 . 54 ) > cluster 12 ( 0 . 03±0 . 37 ) > cluster 21 ( 0 . 01±0 . 83 ) >22 ( −0 . 77±0 . 70 ) . This trend implies that the flow of information follows the path of cluster 11 → 13 → 12 → 21 →22 . Using the same measure of information hierarchy , we found that motor neurons belonging to cluster 21 were located in an earlier processing phase of the information hierarchy than the motor neurons of cluster 22 . The mean value of this parameter for motor neurons of cluster 21 was 0 . 15±0 . 66 , whereas the mean value for the neurons belonging to cluster 22 was 0 . 23±0 . 67 . The value of this parameter also tended to grow as the location of a motor neuron moved posteriorly ( Figure S3 ) , supporting our claim that posterior motor neurons are located at an earlier stage of information processing than anterior motor neurons . From the inward/outward synaptic ratios and the directionality of information flow between clusters , it is plausible to suggest that information flow among the structural clusters identified in this study occurs as follows: ( 1 ) chemosensation: 11 → 12 → head movement for changing direction , 11→ 12 → 21 → 22 → body movement; ( 2 ) mechanosensation: 21 → 22 → body movement . To summarize , the structural clusters indentified in this study appear to serve as a cohesive sub-module for information processing at various stages .
C . elegans is the only organism in which all synapses in the nervous system have been anatomically elucidated . Numerous studies have used this information to investigate how neuronal connections are related to their functions . However , few attempts have been made to identify structurally meaningful clusters by considering the complete wiring diagram of synaptic connections without any prior knowledge or other bias . Analysis of the C . elegans connectome revealed the existence of 5 topological clusters , including a body-spanning cluster , on the individual neuronal level , each of which corresponds to experimentally identified functional circuits . The hierarchical relationships between the five clusters define the systemic cooperation ( e . g . , mechanosensation , chemosensation , and navigation ) between structurally segregated biological circuits toward higher-order complex behaviors . This study explicitly shows structural substrates of functional systems in a micro-scale connectome , which may provide experimentalists with possible predictions for functions of novel circuits in the C . elegans connectome . What is the significance of the existence of distinct structural clusters in the C . elegans connectome ? We show that the nervous system of the nematode , though seemingly simple , is organized into distinct functional modules . A ganglion contains neurons belonging to distinct clusters , suggesting that a ganglion is a simple collection of neurons with their somata lying near each other but also with different functional roles . Thus , synaptic connections make a greater contribution to the biological function of the C . elegans connectome than does the physical location of neuronal cell bodies . We found that each cluster identified through topological clustering exhibited close relationships with its function in neural circuits , supporting our speculation that clustering analysis would be helpful in elucidating the functions of unidentified neurons . Supporting this idea , previous findings of neuronal ablation experiments are consistent with our clustering data . Most command interneurons , except for AVE , are included in cluster 21 ( Figure 4B ) . Cluster 21 also contains the mechanosensory neurons ALM and PLM ( Figure 4B ) . If ALM and PLM neurons are ablated , the worms do not respond to anterior and posterior body touch , respectively . Cluster analysis suggests that the command interneurons contained in cluster 21 are involved in mechanosensation . Consistent with this conclusion , when AVD or PVC neurons were ablated , the worms could not sense anterior or posterior body touch , respectively [40] , [42] . Among the command interneurons , only AVE neurons belonged to cluster 12 , which is consistent with the previous finding that ablation of the AVE pair alone did not result in any locomotion defect [42] . It is possible that , unlike other command interneurons , AVE neurons are involved in connecting chemosensory signals to motor circuits . Using computational output , it is possible to make important predictions about the roles of neurons whose functions have not yet been examined or elucidated . For example , because sensory neurons in cluster 11 are experimentally known to be involved in chemosensation while sensory neurons in cluster 12 are involved in mechanosensation , we can hypothesize that unknown neurons , such as ADA neurons in cluster 11 and IL2 neurons in cluster 12 , may be involved in chemosensation and mechanosensation , respectively . These hypotheses can be experimentally examined . The approach we employed in this study can be extended to other more complex organisms and represents a strong methodology for determining the functional properties of the connectome of other animals . In addition , it will be feasible to test hypotheses based on our information flow using optogenetic methods and neural imaging [46] . For example , after expressing and activating channel rhodopsin proteins in motor neurons belonging to cluster 21 , which are mostly located in the posterior region of the body , one could examine whether neural information can be transmitted to the neurons in cluster 22 . This kind of approach may help to dissect the mechanism of locomotion in more detail . Ablation experiments could also be employed in addition to the inhibitory method using halorhodopsin . Our findings must be interpreted in light of the limitations of this study . Because we lack knowledge on circuit-level information processing in C . elegans neuronal function at present , further validation based on biological experiments is necessary to confirm our findings and to build detailed computational methods for better predictions . Although we analyzed a recent version of the connectome as a directed weighted network , derivation of a more appropriate adjacency matrix of the connectome remains a goal for future theoretical studies . Because the C . elegans connectome contains distinct types of chemical synapses with excitatory or inhibitory synaptic effects , the development of a plausible framework for estimating the correct numbers for each element of the adjacency matrix will be required . | Caenorhabditis elegans ( C . elegans ) is a tiny worm whose neuronal network is fully revealed . Since the modular organization in a network of individual neurons interwoven through synapses is not yet fully explored owing to incredibly complex connectivity architecture , this study is designed to investigate hierarchically organized modules in this complete wiring diagram ( connectome ) of this worm . We used the modularity-based community detection algorithm and found that C . elegans had 5 anatomical clusters in the C . elegans connectome , which corresponded to experimentally-identified functional circuits . We found that the hierarchical organization of the 5 clusters explains the systemic cooperation including mechanosensation , chemosensation , and navigation that occurs among the structurally-segregated biological circuits to produce higher-order complex behaviors . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"neuroscience/behavioral",
"neuroscience"
] | 2011 | Topological Cluster Analysis Reveals the Systemic Organization of the Caenorhabditis elegans Connectome |
Using a DNA polymerase to record intracellular calcium levels has been proposed as a novel neural recording technique , promising massive-scale , single-cell resolution monitoring of large portions of the brain . This technique relies on local storage of neural activity in strands of DNA , followed by offline analysis of that DNA . In simple implementations of this scheme , the time when each nucleotide was written cannot be determined directly by post-hoc DNA sequencing; the timing data must be estimated instead . Here , we use a Dynamic Time Warping-based algorithm to perform this estimation , exploiting correlations between neural activity and observed experimental variables to translate DNA-based signals to an estimate of neural activity over time . This algorithm improves the parallelizability of traditional Dynamic Time Warping , allowing several-fold increases in computation speed . The algorithm also provides a solution to several critical problems with the molecular recording paradigm: determining recording start times and coping with DNA polymerase pausing . The algorithm can generally locate DNA-based records to within <10% of a recording window , allowing for the estimation of unobserved incorporation times and latent neural tunings . We apply our technique to an in silico motor control neuroscience experiment , using the algorithm to estimate both timings of DNA-based data and the directional tuning of motor cortical cells during a center-out reaching task . We also use this algorithm to explore the impact of polymerase characteristics on system performance , determining the precision of a molecular recorder as a function of its kinetic and error-generating properties . We find useful ranges of properties for DNA polymerase-based recorders , providing guidance for future protein engineering attempts . This work demonstrates a useful general extension to dynamic alignment algorithms , as well as direct applications of that extension toward the development of molecular recorders , providing a necessary stepping stone for future biological work .
Our algorithm solves a problem central to interpreting molecular recorder output in the context of neural recording: it aligns a single DNA-based record to an estimate of neural activity . We evaluate the local likelihood of each nucleotide being written at any time within some recording window given some assumed neural and DNAP properties . Then , using a dynamic programming-based technique , we attempt to find a global alignment given the local likelihoods and a prior defined by the DNAP kinetics . This algorithm is similar in structure to Dynamic Time Warping , utilizing a modified step pattern that reflects certain biological realities ( See Algorithm Methods , S1 Fig ) . The step pattern limits the possible search space by enforcing these constraints: 1 ) nucleotides cannot be aligned to the same time point , 2 ) nucleotides can only be aligned to one time point , and 3 ) there can be a variable amount of time between incorporation of two adjacent nucleotides . We weight the potential options from this step pattern so that alignments made more likely by DNAP kinetics are favored . Notably , this approach enables significant algorithm parallelism , emerging from the constraint that nucleotides can only be aligned to one time point . As there are no dependencies between possible alignments of a given nucleotide , we can calculate the costs of all possible alignments of a given nucleotide concurrently . In order to demonstrate the utility of this algorithm , we apply our technique to simulated output of molecular recorders ( Fig 1A ) , demonstrating various aspects of algorithm performance as well as exploring the ability of DNAPs to encode neural information . The general experimental pipeline consists of four parts: ( 1 ) simulation of a molecular recording experiment ( Fig 1B and 1C ) , ( 2 ) alignment of single recorder outputs to a set of time-indexed expected DNAP error rates , which represent potential neural tunings to observed experimental covariates , ( 3 ) selection of a template that best matches the molecular recorder output ( Fig 1D ) , and ( 4 ) inference of neural parameters using time-aligned DNA-based signals ( see Methods ) . We simulate a biologically-inspired generative model with several parts: ( 1 ) an explicit parameterized model of how neural activity either depends on a stimulus or results in observed behavior ( Neural Tuning ) , ( 2 ) how this neural activity modulates DNAP error rate , via Ca2+ concentration or other mechanisms ( DNAP Tuning ) , and ( 3 ) a probabilistic description of DNAP kinetic properties , e . g . incorporation rate and pausing ( DNAP Kinetics ) . This generative model can be parametrized using existing knowledge about neural and polymerase properties where known . In this paper , we use DNAPs with optimistic DNAP error tuning , i . e . maximum error rates higher than many DNAPs with incorporation rates suitable for recording , but with otherwise-realistic properties [19–21] . We also assume knowledge of these system characteristics ( apart from neural tuning ) in order to parametrize the alignment algorithm . Given simulated DNA output and a time-varying input to the system , we iterate over potential neural tunings to find a tuning that provides an alignment most consistent with the observed DNA-based signal . We then use this maximum a posteriori alignment to generate a time-indexed DNA signal , and use this signal to infer neural parameters . We evaluate algorithm performance both by accuracy of timing estimation , i . e . how many seconds estimated incorporation times differ from true incorporation times on average , and accuracy of inferred neural tuning parameters , i . e . how the estimated behavior of a neuron differs from the true neural behavior . Specifically , to evaluate accuracy of timing estimates , we examine the root-mean square deviation ( RMSD ) between the estimated timings and the true incorporation times for a given alignment . There is a highly non-linear relationship between alignment “success” and timing accuracy , as nearby alignments do not necessarily have similar likelihoods . Thus , we provide both a mean and median value for timing accuracy when those values differ by a large amount . To evaluate tuning accuracy , we estimate tuning parameters from the aligned DNA data and examine the distance between the algorithm-estimated parameters and those derived directly from the recorded neural data , which we treat as ground-truth for these studies .
Before exploring algorithm applications , it is worth exploring the performance implications of this approach . It bears mentioning again that , while they do not calculate the same cost function , our algorithm and traditional DTW are closely related; both are dynamic programming algorithms with effective worst-case complexity of O ( NT ) where N and T are the lengths of the two inputs being aligned . As we have mentioned , our algorithm has significant differences in implementation that allow it to be substantially parallelized; this allows for substantial performance increases using modern computing devices ( See Algorithm Methods ) . While a naïve implementation of our algorithm performs more slowly than traditional DTW for a given set of inputs , parallelized implementations substantially outperform traditional DTW ( S1 Fig ) . We observe up to a 16x speedup over traditional DTW when using a GPU-based implementation of our algorithm on a personal computer , and up to a 5x speedup when using a CPU-based implementation . The feasibility of a “ticker tape” DNA-based recording scheme depends heavily on the properties of the DNAP used . For instance , the length of records ( in base pairs ) influences how much information is contained about neural activity , and thus impacts algorithm performance . Similarly , the speed , pausing , and fidelity properties of the DNAP used influence the information about neural activity contained in a DNA-based record [7] . Here , we look to determine the effect of these properties on the accuracy of our algorithm , and thus the expected performance of a molecular recording setup . Determining these effects allow us to form guidelines as to what kinds of DNAPs would be required for successful recording and alignment . We use an entirely-simulated experiment here , i . e . we fully know the tuning linking stimulus to neural activity . This allows us to isolate the effects of DNAP properties on alignment from the effects of inaccurate neural activity estimates . We simulate a neuron with a linear response to an artificial stimulus; we deliver random levels of stimulus in 5s blocks over the course of 2000s ( ~30 minute recording window ) , and simulate the neuron’s spiking activity and intracellular calcium . We then simulate the output of a molecular recording system during that time period . We then align the molecular recorder output to the true stimulus signal . Using this simulation , we can focus on error induced by the DNAP and alignment algorithm in isolation . We aim to estimate nucleotide incorporation timings , as well as the strength of the neuron’s tuning to the stimulus , i . e . the slope of the neuron’s tuning curve . The best alignments possible under this scheme have timing error up to the size of the stimulus features ( 5s ) ; alignments with timing error less than this are generally considered to be accurate . Error with respect to tuning parameter is presented as a proportion of the true parameter . Except for the DNAP parameter being varied , the simulated DNAPs are identical ( ~100 Hz , mean pause duration of 2s; see Methods ) . As record length increases , finding a randomly generated pattern that resembles the record becomes less likely , and alignment to a unique site should become easier . However , from a biological perspective , generating longer sequences may be more difficult , requiring polymerases with specialized properties , e . g . high processivity , high activity , or strand-displacement activity . Thus , it is useful to know minimal record lengths for successful alignment . When we increase record length in our simulations , we indeed find a resulting decreasing timing error . Generally , we find that records with length longer than 2 . 5K basepairs align with <5s median timing error ( Fig 2A and 2B ) . Interestingly , we find that slope estimation is relatively constant regardless of record length , suggesting that , while record length is crucial to timing estimation , information about neural tuning in the record is not necessarily absent in shorter records ( Fig 2C ) . DNAP speed effectively changes the sampling rate of our system; if we have a slow DNAP , we can record for longer periods of time for a given strand length , but also record less information about any given interval . If we are interested in longer time-scale phenomena ( e . g . environmental sensing , medical diagnostics ) [22] , we may wish to use slow DNAPs . However , due to the low sampling rate , we may not be able to recover useful information about timing and tuning in a neural paradigm . In our simulated stimulation paradigm , we find that slower DNAPs in fact increase timing accuracy ( Fig 2D ) . However , median timing error stays relatively constant as speed decreases , implying that slow DNAPs simply decrease the amount of extreme timing errors we observe ( Fig 2E ) . This runs parallel to our observations about record length; aligning to a longer time-indexed template is easier than aligning to a short one . However , our accuracy in determining tuning parameters decreases as we use slower DNAPs ( Fig 2F ) . This indicates that we should , in general , be using fast DNAPs if we are interested in recovering tunings [19] . Meanwhile , slower DNAPs can provide longer records for a given strand length at the expense of diluting the information they carry about underlying phenomena . Another property of DNAPs that can affect the quality of recordings is the transfer function relating analyte ( e . g . calcium ) concentration to error rate , f ( · ) . We have modeled f ( · ) as a sigmoid with three parameters: f ( C ) =Rmax⋅11+exp[b ( C−C0 ) ] ( 1 . 1 ) where C0 denotes the [Ca2+] that leads to half-maximum error rate , b denotes the steepness of the response curve , and Rmax denotes the maximum error rate of the DNAP . When selecting ( or engineering ) DNAPs to record with , we will need to optimize over these parameters . Here , we analyze DNAPs with varying transfer function slopes b , i . e . varying sensitivities to [Ca2+] , ranging from step-like DNAPs to DNAPs with a wide dynamic range . We find that DNAPs with moderate sensitivities to [Ca2+] provide the most accurate timings , while both step-like and overly shallow transfer functions decrease alignment accuracy ( Fig 2G and 2H ) . We find similar results for parameter estimation ( Fig 2I ) , where appropriately-sloped DNAP tunings provide better estimates of neural parameters than DNAPs that are either too insensitive ( low |b| ) or too step-like ( high |b| ) with respect to [Ca2+] . This adds evidence to an assumption many investigating molecular recording techniques have been working under: DNAPs will have to be tailored in order to achieve optimal recording of even simple signals . We are also interested in how the maximum error rate Rmax affects alignment accuracy . This is of particular interest from a biological perspective: many natural DNAPs with incorporation rates suitable for high-resolution recording have low error rates . It is useful to understand what minimal error rates would be feasible for molecular recorders , as well as examine system performance as Rmax scales . Here , we consider DNAPs that have near-zero error rates at low [Ca2+] , and increase to some maximum error rate Rmax under high [Ca2+] conditions . We find that alignment accuracy increases as maximum error rate increases ( Fig 2J and 2K ) , as expected . Interestingly , we find that parameter estimation is relatively insensitive to Rmax . Again , this seems to suggest that while timing accuracy tends to degrade with unfavorable DNAP parameters , molecular recorder output tends to retain information about underlying neural tuning . Here , we demonstrate the feasibility of molecular recorders in a conventional neuroscience experimental paradigm . We analyze single-unit neural data recorded from M1 and pre-motor cortex during a center-out reaching task in a rhesus macaque , estimating the preferred movement directions of recorded neurons ( data obtained from the DREAM reaching experiment database , see Flint 2012 for details [23–25] ) . We use the recorded spikes as the basis for simulated calcium transients and molecular recorder output . We also generate a set of estimates of neural activity from the kinematic data recorded during the task , with estimates representing velocity-tuned neurons with preferred directions distributed uniformly on [0 , 2π] . Here , we use eight activity estimates as alignment templates . We apply our alignment algorithm to this data , aligning the molecular recorder output to each of the estimates , then selecting the maximum-likelihood alignment . The result , an estimated mapping of nucleotides to time , allows us to generate tuning curves for the recorded neurons . From this , we can estimate neural tuning parameters and infer how neural activity is modulated with respect to the recorded kinematics ( details in Methods ) . The alignments here encounter alignment- and DNAP-based error , as in the previous section , but also encounter biology-based error when estimating neural activity from kinematic data . Thus , these experiments serve as an estimate of molecular recorder performance in a real-world scenario . Using a plausible set of DNAP parameters ( ~100 Hz incorporation rate , mean pause duration of 2s , ~17% of time spent paused; see Methods for further details ) , we find that we are generally able to recover rough timing estimates and accurate tuning parameters from the simulated molecular recording experiment . As an initial demonstration , we examine several neurons that exhibit high firing rates and significant directional tuning ( Fig 3A ) . Under these conditions , we are able to estimate nucleotide timings to within an average of ~15s ( 95% confidence intervals for average trial RMSD: [10 . 0 , 16 . 5] , [12 . 1 , 20 . 3] , and [14 . 8 , 22 . 5] seconds , Fig 3B ) . While timing accuracy is lower than desired , particularly for experiments that require sub-second precision using current techniques , these alignments still allow us to generate the estimated neural tuning direction θ* with error of ~10% ( average errors of 0 . 5 , 0 . 3 , and 0 . 3 radians , Fig 3C ) . Median timings are substantially better than average timings across the board ( 95% confidence intervals for median trial RMSD: [3 . 8 , 7 . 2] , [3 . 1 , 8 . 7] , and [6 . 5 , 13 . 7] seconds ) . Some of the error we encounter when generating alignment estimates may stem from our discrete parametrization of neural tunings . That is , we may not provide an estimate of neural activity similar enough to the true activity in order to generate accurate alignments . We can examine the contributions of this effect to algorithm accuracy by supplying a neural activity estimate generated using the neural tuning estimated from electrophysiology data , the best possible estimate we can provide given a particular model . Indeed , if we supply a neural activity estimate generated using the ground-truth neural preferred direction in our motor control experiment ( rather than the 8 naïve preferred directions ) , we substantially reduce both timing error and error in θ* ( S2 Fig ) . While we do not know the true preferred direction a priori and this kind of analysis could not be performed in practice , this suggests that a large portion of observed error can be attributed to the discrete parametrization of the search space . Increasing the resolution of the search space should improve alignment accuracy at the expense of execution time . We apply our algorithm to each neuron in the dataset , examining aggregate performance over a population of recorded neurons . We find that the technique has middling performance on the whole dataset , only able to estimate timings to within 24s for 12% of neurons recorded ( S3 Fig ) . If we limit the set of analyzed neurons to those that have substantial reach-modulated activity ( model pseudo-R2 > 0 . 05 , firing rate λ > 20 spikes/s ) , this improves to 47% . We are able to estimate preferred direction to within ±0 . 2π ( ±36° ) for 39% of the dataset; this improves to 59% of the reach-modulated neurons ( S3 Fig ) . While this filtering does not explain all observed error , it is useful when reconciling the results for individual neurons in Fig 3 with the larger dataset . This improvement upon filtering for active , well-modeled neurons demonstrates two things: 1 ) this method performs poorly on sparse-firing neurons , and 2 ) this method performs poorly on neurons that are not well-described by the set of models we consider . Both of these shortcomings are as expected given the algorithm . The former can be addressed by evaluating average neural activity represented by a DNA-based record , which can be done in a naïve , model-free manner . The latter , an inability to align signals that we cannot already model accurately , remains a shortcoming of this approach when attempting the interpretation of molecular recorder output . We also analyze recording systems with a hypothetical DNAP that exhibits no pausing , but is otherwise identical to the previous DNAPs ( see Methods ) . When examining the same neurons as above , we find drastically decreased timing errors ( RMSD 95% CIs of [0 . 17 , 0 . 18] , [0 . 31 , 0 . 39] , and [0 . 47 , 3 . 0] seconds ) and parameter estimation errors ( average errors of 0 . 1 , 0 . 2 , and -0 . 04 radians , S4 Fig ) . Using these highly optimized DNAPs , we approach the timing resolution that would seem to be useful for high-precision neuroscience experiments , and retain high-accuracy prediction of neural tunings . A conclusion from this analysis is that much of the error we observe with our technique resolves when DNAPs behave more regularly . These results are of particular interest to us because of their biological implications: DNAP pausing generally has both DNAP-based and sequence-dependent components , and can be ablated using sequence context , chemical , or temperature-based means [19 , 26 , 27] . This significant improvement in both timing accuracy and parameter estimation suggest that decreasing DNAP pausing through these or other methods could be a useful approach to improve the accuracy of molecular recording systems . We observe that errors in tuning parameter estimation in our simulated reaching experiments are not always normally distributed; rather , in a number of neurons , there appear to be several preferred directions that alignments converge upon , including peaks at a neuron’s anti-tuned direction ( Fig 3C ) . This effect persists , although less prominently , when using a non-pausing DNAP ( S4 Fig ) . This is useful to consider given the underlying center-out task in our experiment , where subjects reach in a direction then immediately make a reach back to the center , i . e . the opposite direction of the initial reach . It seemed possible that pathologic alignments could arise from this repetitive temporal structure , where alignments to tuned and anti-tuned templates are effectively identical save for a time-lag . Disrupting this structure through appropriate experimental design could lead to improved accuracy . We generated a dataset composed of shuffled 2-second-long patches of neural and kinetics data such that the temporal structure of the original dataset was disrupted . We find that shuffling the data can both reduce selection of anti-tuned preferred directions ( Fig 4A and 4B ) , as well as decrease overall tuning estimation error ( Fig 4C ) . However , it is important to note that the shuffling scheme we describe here does not improve alignment for all neurons , and can even disrupt alignment of neurons that are otherwise predicted correctly ( S5 Fig ) . While this argues against naïve shuffling as a universal strategy , it further demonstrates the effect of an experiment’s temporal structure on alignment accuracy . These findings suggest that experimental design cognizant of alignment-based analysis can improve robustness to pathologic alignments , and thus the feasibility of molecular recording-type experiments .
There are many ways in which existing DNAPs already satisfy the requirements necessary for a single-strand biological recorder , e . g . processivity , speed , calcium-sensitive error rates , and pausing kinetics [19 , 26 , 29] . The one property that we have not observed in DNAPs is a calcium-sensitive error rate at physiological concentrations [20] . Further , natural DNAPs tend to be either fast or error-prone , but not generally both; the highest error rates we see in high-incorporation-rate DNAPs are at the low end of what we simulate here [21 , 30] . In order to develop practical molecular recorders , we will both need to understand how to substantially increase DNAP error rates in processive , high-speed DNAPs , as well as develop a scheme to make DNAP error rates calcium-sensitive at physiologically relevant scales . Alternatively , schemes that do not rely on calcium-tuned error rates , but rather modulate other DNAP properties via calcium , may provide an easier way forward . While many caveats apply to this work , and to the prospect of molecular recorders in general , the results described here are helpful on a number of fronts . On a technical side , we describe a DTW-class algorithm that applies generally to point processes with variable temporal indexing . The algorithm is designed to allow probabilistic interpretation of its output , and can be used to find maximum a posteriori alignments to a set of known templates . We provide a highly-parallelized implementation of this algorithm which leverages advances in asynchronous computing techniques . With respect to molecular recorders , we provide a framework for interpretation of recorder output in the face of uncertain recording times . We also provide guidance to the ongoing research that looks to engineer DNAPs for this kind of recording . Perhaps most importantly , we have shown that , should a DNAP with certain properties be developed , we can provide temporal indexing to its output and capture neural behaviors using a molecular recording approach . While this is purely a simulation study , our work sets constraints and goals for the development of DNAPs for massive-scale neural data recording , and outlines experimental scenarios for their successful use .
This technique is intended to align a DNA-based recording with no temporal indexing to a longer , time-indexed estimation of calcium activity , a template . It assumes the DNA sequence as a binary “error”/”no error” code , then assesses the similarity of that sequence to a discrete-time continuously-valued estimate of neural activity , the template , via alignment . We use a novel DTW-class algorithm to perform this alignment , incorporating beliefs about DNAP kinetics to limit the space of potential actions . | This work demonstrates a necessary computational tool for the development and implementation of molecular recorders , a promising potential technique for massive-scale neuroscience . Molecular recorders use proteins to encode levels of a substance we want to measure ( e . g . calcium in neural applications ) as detectable changes in a linear cellular structure , e . g . misincorporations in a strand of DNA , or fluorescent proteins traveling down a microtubule . This encoding represents levels of the measured substance over time , much like a ticker tape represents information linearly on a strip of paper . The unique intracellular nature of this approach promises a significant scaling advantage over current techniques . The molecular recording approach suffers a particular drawback involving timing: unlike most methods of recording signals , in simple molecular recording systems we do not observe when each data point was recorded . This timing information is almost always required in order to make associations between our recorded data and the rest of the experiment . In this work , we propose a method to estimate the timing of these data points using easily-observable experimental measurements . We demonstrate the application of this method in a simulated neuroscience paradigm , investigate the effect of experimental design on this method , and determine protein properties that would be desirable in molecular recorders . These findings are useful both as a computational proof-of-concept , and as guidelines for current efforts to engineer proteins for molecular recording . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"molecular",
"neuroscience",
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"applied",
"mathematics",
"membrane",
"potential",
"experimental",
"design",
"electrophysiology",
"neuroscience",
"simulation",
"and",
"modeling",
"algorithms",
"research",
"design... | 2017 | Nucleotide-time alignment for molecular recorders |
Pseudomonas aeruginosa ( P . aeruginosa ) is a major opportunistic human pathogen , causing serious nosocomial infections among immunocompromised patients by multi-determinant virulence and high antibiotic resistance . The CzcR-CzcS signal transduction system in P . aeruginosa is primarily involved in metal detoxification and antibiotic resistance through co-regulating cross-resistance between Zn ( II ) and carbapenem antibiotics . Although the intracellular regulatory pathway is well-established , the mechanism by which extracellular sensor domain of histidine kinase ( HK ) CzcS responds to Zn ( II ) stimulus to trigger downstream signal transduction remains unclear . Here we determined the crystal structure of the CzcS sensor domain ( CzcS SD ) in complex with Zn ( II ) at 1 . 7 Å resolution . This is the first three-dimensional structural view of Zn ( II ) -sensor domain of the two-component system ( TCS ) . The CzcS SD is of α/β-fold in nature , and it senses the Zn ( II ) stimulus at micromole level in a tetrahedral geometry through its symmetry-related residues ( His55 and Asp60 ) on the dimer interface . Though the CzcS SD resembles the PhoQ-DcuS-CitA ( PDC ) superfamily member , it interacts with the effector in a novel domain with the N-terminal α-helices rather than the conserved β-sheets pocket . The dimerization of the N-terminal H1 and H1’ α-helices is of primary importance for the activity of HK CzcS . This study provides preliminary insight into the molecular mechanism of Zn ( II ) sensing and signaling transduction by the HK CzcS , which will be beneficial to understand how the pathogen P . aeruginosa resists to high levels of heavy metals and antimicrobial agents .
Bacteria are extremely versatile that can regulate cellular processes in a sophisticated manner and thereby survive in changing environments . The two-component system ( TCS ) is the predominant strategy for coupling various extracellular stimuli to appropriate cellular responses in microorganisms [1–5] . In Pseudomonas aeruginosa ( P . aeruginosa ) , approximately 130 genes have been identified that encode various types of TCSs [5 , 6] . These regulatory systems enable this organism to ubiquitously exist in diverse environments and to express various virulence factors [7 , 8] . Thus , P . aeruginosa is one of the most prevalent opportunistic pathogens and causes severe hospital-acquired infections among immunocompromised patients [7 , 9 , 10] . It is capable of causing both chronic and acute pulmonary infections in cystic fibrosis ( CF ) patients , ventilator-associated pneumonia , and sepsis in burn patients [8] . Moreover , this pathogenic bacterium possesses intrinsically high levels of resistance to multiple classes of antimicrobial agents , presenting tremendous obstacles for anti-infective therapies [7 , 11] . The CzcR-CzcS TCS in P . aeruginosa is responsible for numerous cellular processes , including Zn ( II ) resistance , carbapenem antibiotic resistance , quorum sensing , and virulence regulation ( Fig 1A ) [12–14] . Under direct stimulation by Zn ( II ) , the histidine kinase ( HK ) CzcS auto-phosphorylates on its conserved histidine residue . It subsequently transmits the phosphoryl group to the conserved aspartate residue in the receiver domain of the response regulator ( RR ) CzcR . The phosphorylated CzcR up-regulates the expression of a metal efflux pump , CzcCBA . It also represses the expression of OprD , a porin that regulates the entry of basic amino acids and carbapenem antibiotics [13 , 14] . This co-regulation between metal detoxification and antibiotic resistance is unusual , and its mechanism will provide significant guidance for the treatments of environmental and clinical issues . In the CzcR-CzcS TCS , the HK CzcS is predicted to be the transmembrane sensor-transmitter , and it contains three functional domains [15] . The highly diverse N-terminal periplasmic sensor domain is arranged between two membrane-spanning segments and is followed by a conserved C-terminal cytoplasmic kinase domain . The stimulus is detected by the periplasmic sensor domain and transmitted across the membrane to the cytoplasmic kinase domain [3 , 16] . Given the pivotal role of the extracellular sensor domain in signal recognition and transduction , we determined the crystal structure of the CzcS sensor domain ( CzcS SD ) in the presence of Zn ( II ) ( referred as CzcS-Zn hereafter ) . Together with the biochemical and in vivo studies , the CzcS SD is identified to bind Zn ( II ) between N-terminal H1 and H1’ α-helices , which is the key first step in Zn ( II ) detoxification and meropenem resistance by HK CzcS . The N-terminal H1 and H1’ α-helices are also shown to play important roles in signal transduction via a series of structure-guided mutagenesis studies . The study reveals the CzcS SD appears to utilize a new mode which is not previously observed for sensor HKs to protect P . aeruginosa from high levels of Zn ( II ) and in parallel meropenem .
The structure of CzcS-Zn complex was solved by single-wavelength anomalous diffraction ( SAD ) using the data collected at the zinc peak wavelength ( 1 . 2823 Å ) and was refined using data collected at a wavelength of 1 . 0000 Å . The Rwork and Rfree are 0 . 210 and 0 . 253 , respectively . The data collection and other refinement statistics are summarized in Table 1 . The structure belongs to the C2 space group and it contains two molecules ( CzcS SD , amino acids 40–166 in the CzcS protein ) per asymmetric unit ( S1A Fig ) . As supported by the clear electron density , residues 40–132 and 135–161 in both molecules are well defined , whereas the C-terminal segment ( residues 162–166 ) and the loop ( residues 133–135 ) that connects the S4 and S5 β-strands are disordered . The tertiary structures of the two molecules are similar ( S1B Fig ) . The root-mean-square deviation ( r . m . s . d . ) between them is 0 . 8Å with 104 pairs of corresponding Cα atoms superimposed . The structural deviations are mainly caused by the tilting of the N-terminal domain ( residues 40–81 ) with respect to the C-terminal domain ( residues 83–161 ) . The structure of CzcS SD is a mixed α/β-fold in nature , which can be divided into two domains ( Fig 1B ) . The N-terminal helix-loop-helix domain is composed of the H1 α-helix ( residues 40–60 ) , the connecting loop ( residues 61–66 ) , and the H2 α-helix ( residues 67–81 ) . It is connected to the C-terminal domain by one residue , Thr82 . The C-terminal domain contains five β-strands ( S1-S5 ) and one 03B1-helix ( H3 ) . The five β-strands form one anti-parallel β-sheet . S5 ( residues 150–161 ) is located in the middle and is flanked by S1 ( residues 83–90 ) and S2 ( residues 97–102 ) on one side and by S4 ( residues 136–147 ) and S3 ( residues 126–132 ) on the other side . The H3 α-helix packs against the anti-parallel β-sheet and is nearly perpendicular to the first two α-helices . A short kink at residues 106 and 107 divides the H3 α-helix into H3a ( residues 103–105 ) and H3b ( residues 108–120 ) . The N-terminal H1 α-helix and the C-terminal S5 β-strand are oriented in the same direction , which connect to the transmembrane helix ( TM1 and TM2 helices ) in the transmembrane domain of HK CzcS . The typical PDC members include the divalent cation sensor PhoQ [18 , 19] , the citrate sensor CitA [20 , 21] , and the C4-dicarboxylate sensor DcuS [22] . Despite of the negligible sequences identities , it reveals that the CzcS SD possesses a similar structural arrangement to those of other members of the PDC superfamily by the structural similarity searches performed with the Dali server program [23] ( S2 Fig ) . The relatively high Z-scores of 5 . 4 , 4 . 0 , and 7 . 0 are yielded in the structure-based alignments of the CzcS SD with PhoQ [18] , CitA [20] , and DcuS [22] , respectively . The sensor domain of CzcS and PhoQ can be largely superimposed in the C-terminal domain with a r . m . s . d . of 2 . 6 Å over 66 corresponding Cα atoms . Similarly , the CitA superimposes onto CzcS over 58 corresponding Cα positions with a r . m . s . d . of 2 . 9 Å , and DcuS superimposes onto CzcS over 54 corresponding Cα positions with a r . m . s . d . of 2 . 6 Å . The distinct difference between the structure of CzcS SD and those of the other PDC superfamily members is the orientation of the N-terminal helix-loop-helix domain ( Fig 1C ) . In the structure of CzcS-Zn , the N-terminal helix-loop-helix domain is tilted away from the central five anti-parallel β-sheet , which may be caused by the Zn ( II ) binding at the H1 and H1’ α-helices . Two Zn ( II ) ions are captured in the structure of CzcS-Zn . One of the Zn ( II ) ions is coordinated with His72 and Asp76 of CzcS molecule A ( S1A Fig , gray ) . It is also coordinated with Asp62 of symmetry-related molecule B and His72 of symmetry-related molecule C ( S3 Fig ) . The P . aeruginosa functions normally in response to extracellular Zn ( II ) with the double mutation of His72 and Asp76 on HK CzcS ( S4 Fig ) . The result shows that this Zn ( II ) binding pattern is physiologically irrelevant and may be caused by crystallographic packing . The other Zn ( II ) is identified to be functional relevant and is associated with the second CzcS molecule ( S1A Fig , green ) shown in the asymmetric unit . Double mutation of coordinated residues ( His55 and Asp60 ) of second Zn ( II ) on HK CzcS causes severely defects on the regulation of Zn ( II ) resistance in P . aeruginosa ( S4 Fig ) . In the symmetry operation , this CzcS molecule can form a homodimer ( Fig 1B ) , and the Zn ( II ) is exclusively buried between the central parallel H1 and H1’ α-helices which constitute the dimer interface ( Fig 2A ) . The H1 and H1’ α-helices are surrounded by multiple solvent water molecules , which facilitate the Zn ( II ) access to the active site . A distorted tetrahedral geometry is adopted by Zn ( II ) to coordinate with the symmetric ligands ( His55/Asp60 and His55’/Asp60 ) from the H1 and H1’ α-helices , respectively ( Fig 2B ) . The His55 and His55’ residues interact with the Zn ( II ) through their Nε2 nitrogen atoms , and Asp60 and Asp60’ residues contact Zn ( II ) through the Oδ2 atoms of their carboxylate side-chain ( Fig 2B ) . The bond distances of the coordination center are 1 . 92 Å-2 . 17 Å with the bond angles ranging from 102 . 0° to 127 . 6° . In each monomer , the Oδ2 atom of Asp59 residue makes a hydrogen-bond interaction with the Nδ1 nitrogen atom of the His55 residue . Additionally , the O atom of His55 forms hydrogen-bond with the N atoms of Asp59 and Asp60 , respectively ( Fig 2B ) . These second shell interactions , particularly the carboxylate side-chain with the histidine ligand , are thought to play an important role in the stability of the coordination structure [24] . In view of the co-regulation of cross-resistance between metal ions and carbapenem antibiotics by CzcR-CzcS TCS , the physiological importance of the His55 and Asp60 residues in P . aeruginosa is investigated by using the Zn ( II ) and meropenem ( MEPM ) antibiotic tolerance assay ( Fig 2C ) . All the strains keep consistent growth state on the LB solid medium in the absence of Zn ( II ) , and their growing status are not influenced until the concentration of Zn ( II ) reaches 0 . 5 mM . However , when a higher concentration of Zn ( II ) ( 2 . 5 mM ) are supplied , the P . aeruginosa PAO1 strain demonstrates an obvious growth advantage over the czcS-deficient strain . The abolished metal resistance of the czcS-deficient strain is restored to wild type levels by complementation with a plasmid ( pAK1900 ) carrying the czcS gene . When His55 and Asp60 are substituted by the amino acids which can’t coordinate with Zn ( II ) , the mutant strains ( H55A , D60A , and H55R ) dramatically attenuate their abilities in the Zn ( II ) detoxification for the destruction of Zn ( II ) binding site . Remarkably , the losses in responsiveness to Zn ( II ) of H55A and H55R mutant strains are similar in degree to the non-responsiveness of the czcS-deficient strain . Additionally , the aforementioned mutant strains and the czcS-deficient strain lose their Zn ( II ) -inducible resistance to the MEPM antibiotic . The mutagenesis analyses corroborate the crucial role of the His55 and Asp60 in Zn ( II ) sensing . Intriguingly , markedly different phenotypes are observed when the His55 and Asp60 are replaced with the coordinated cysteine residues ( Fig 2C ) . The H55C mutant partially preserves the ability in Zn ( II ) detoxification and in parallel meropenem resistance . By contrast , the D60C mutant shows equivalent responsiveness to Zn ( II ) and MEPM to that of the wild type PAO1 . The Co ( II ) has similar radius to Zn ( II ) , and it can bind to Cys2His2 coordination site in a tetrahedral geometry as well [25] . The wild type P . aeruginosa is blind to Co ( II ) that none Co ( II ) -inducible resistance to the MEPM antibiotic can be observed ( Fig 3 ) . With the mutation of residues ( Asp60 and Asp60’ ) on the N-terminal H1 and H1’ α-helices , the D60C mutant strain displays Co ( II ) -inducible resistance to the MEPM antibiotic ( Fig 3 ) . Although the increased antibiotic resistance induced by Co ( II ) is not as strong as that of Zn ( II ) , this experiment indicates that the binding of Co ( II ) between the N-terminal H1 and H1’ α-helices can also regulate the downstream signaling transduction in CzcR-CzcS TCS . The linker region connects the H1 and H1’ α-helices to the transmembrane helices . It plays an important role in the process of signaling transduction from extracellular sensor domain to the transmembrane region . The cysteine substitution scanning is performed in the linker region on the basis of the H55A mutant ( Fig 4A ) . As demonstrated above , the H55A mutant loses its intrinsic resistance to Zn ( II ) and MEPM as the czcS-deficient strain due to the destruction of Zn ( II ) binding site . In conjunction with the mutation L38C , the L38C H55A mutant restores the responsiveness to Zn ( II ) stimulus . This mutant strain can survive on the LB medium with high concentrations of Zn ( II ) and shows Zn ( II ) -inducible resistance to MEPM ( Fig 4B ) . The experiment indicates that the strain with cysteine substitution in the linker region instead of that at the Zn ( II ) binding site can also sense and transmit the Zn ( II ) signal as well as wild type P . aeruginosa . The chromogenic indicator 4- ( 2-Pyridylazo ) resorcinol ( PAR ) is reported to form both 1:1 and 2:1 complexes with Zn ( II ) with stepwise affinity constants of 7 . 7×106 and 5 . 0×105 M-1 , respectively ( at pH 7 . 4 , 0 . 15 M KCl , 22°C ) [26 , 27] . It has been widely used to determine the dissociation equilibrium constants of protein-Zn ( II ) complex in the range of nanomolar to picomolar [28 , 29] . With the addition of Zn ( II ) to the PAR solution , the formative PAR-Zn and PAR2-Zn complex will cause an intense absorbance at 500 nm [29] . The absorption bands of PAR and Zn ( II ) complex at 500 nm are reduced by the addition of wild type and mutant CzcS SD ( CzcS SD H55C , CzcS SD D60C , and CzcS SD L38C H55A ) ( S5 Fig ) . It indicates that the wild type CzcS SD and aforementioned mutants have the ability to compete with PAR for binding Zn ( II ) . The representative titration spectrums are displayed in S6 Fig for the PAR with Zn ( II ) under the competition of wild type and mutant CzcS SD . The titration data at 500 nm and fitting binding isotherms of which are inserted in corresponding titration spectrums ( S6 Fig ) . The dissociation constants determined by Dynafit software [30] for Zn ( II ) with wild type CzcS SD , CzcS SD H55C , CzcS SD D60C , and CzcS SD L38C H55A are 1 . 7 ( ±0 . 2 ) ×10−6 M , 8 . 5 ( ±0 . 4 ) ×10−7 M , 5 . 7 ( ±0 . 3 ) ×10−8 M , and 9 . 4 ( ±0 . 8 ) ×10−9 M , respectively . By using one-site fitting model , the coefficient of variation ( CV ) for each equilibrium binding experiments is approximately 10% regardless of whether the wild type or mutant CzcS SD is monitored ( S1 Table ) . It indicates that the model for data fitting provides a good description of the available data . The Zn ( II ) induced dimerization of CzcS SD was analyzed by the chemical crosslinking experiments ( S7A Fig ) with bis[sulfosuccinimidyl] suberate ( BS3 ) as the primary amine reactive crosslinker [31–33] . Under denatured electrophoresis conditions , the CzcS SD primarily migrates at the position with the molecular weight of monomer , and negligible proportion of dimer is observed when it is treated by excess BS3 crosslinker ( S7A Fig ) . It means that the CzcS SD without Zn ( II ) is mainly existed as monomer in solution that leads the poor efficiency of intermolecular crosslinking reaction . The efficiency of intermolecular crosslinking is significantly increased when Zn ( II ) is loaded into CzcS SD ( S7A Fig ) . With the Zn ( II ) binding at dimer interface ( Fig 2A ) , the dimer produced by the intermolecular crosslinking reaction is analyzed to be as high as 33 ( ±5 ) % of the original sample by using ImageJ analysis [34] . Other divalent cations such as Mg ( II ) , Co ( II ) , and Mn ( II ) are performed in the chemical crosslinking experiments , too . None of these divalent cations can induce crosslinked dimerization of wild type CzcS SD , which shows that Zn ( II ) is specific for this crosslinking ( S7A Fig ) . However , the Co ( II ) show its ability in inducing the crosslinked dimerization of mutant CzcS SD D60C ( S7B Fig ) . The proportion of dimer induced by Co ( II ) is less than that of Zn ( II ) ( S7C Fig ) . This may be caused by the different coordination geometry preferred by Co ( II ) and Zn ( II ) . As the data from Protein Data Bank ( PDB ) and Cambridge Structural Database ( CSD ) , the octahedral geometry is preferred by Co ( II ) coordination and tetrahedral geometry is more preferred by Zn ( II ) coordination . It is well known that proline residue will distort the regular structure of helices by introducing a kink between the segments adjacent to it [35–37] . In this study , we want to characterize whether the distortion of H1 and H1’ α-helices has any influence on the function of HK CzcS by the introduction of proline residue . Besides the residues adjacent to the C-terminal Zn ( II ) binding site , other residues along the H1 and H1’ α-helices are chosen to do the proline substitutions . The residues Gln52 , Leu50 , Leu48 , Asn45 , Arg43 , and Arg41 arranged with almost equal interval are replaced by proline residues ( Fig 5A ) . All these single proline substitutional mutants grow well in the low concentration of Zn ( II ) ( 0 . 5 mM ) and respond to MEPM equivalently to that of wild type P . aeruginosa . However , most of the mutants ( L50P , N45P , R43P , and R41P ) display varying degrees of impairments in their resistance to higher concentration of Zn ( II ) ( 2 . 5 mM ) and the Zn ( II ) -inducible cross-resistance to MEPM ( Fig 5B ) . Unlike the proline substitutions , other non-conserved mutations ( S2 Table ) of these residues ( Arg41 , Arg43 , and Arg45 ) do not profoundly affect the signaling response . We also find that some proline substitutions still play full functions in Zn ( II ) induced metal detoxification and MEPM antibiotic resistance , such as Q52P and L48P ( Fig 5B ) . Further , we also do some double proline substitutions along the H1 and H1’ α-helices ( S8 Fig ) . All the double mutants ( R41P L48P , R43P L48P , N45P L48P , R41P N45P , R43P N45P , and R41P R43P ) totally lose the functions in Zn ( II ) induced metal detoxification and MEPM antibiotic resistance .
TCSs are frequently used by bacteria to adapt to the dynamic environments by coping with external stimuli [1–5] . Since their first discovery approximately 25 years ago , TCSs have been extensively identified in microorganisms . Although more and more crystal structures of HKs have been determined in the last few years [38 , 39] , the crystallographic analyses of Zn ( II ) -binding senor domains of HKs have not been reported to date . What’s more , the mechanisms by which extracellular stimuli are transduced from the sensor domain to the intracellular kinase domains are one of the least understood aspects of TCS response [38] . With growing concerns about co-regulation between heavy metals and antibiotic resistance [13] , the CzcR-CzcS TCS of the pathological bacterium P . aeruginosa is an excellent candidate to study on . Here , we present a high-resolution structure of the CzcS SD in complex with its cognate ligand , Zn ( II ) . The characterized Zn ( II ) -bound CzcS SD is demonstrated as a functional dimer with its central parallel bundle formed by the H1 and H1’ α-helices ( Fig 2A ) . The subunit of the CzcS SD reveals characteristic PDC folds with N-terminal helix-loop-helix domain that leads into the central five anti-parallel β-sheet scaffold ( Fig 1C ) . The N-terminal and C-terminal ends of the CzcS SD are orient in parallel , which allows communication of the connected transmembrane segments with the structural character of dimeric four-helical bundles [40] . The topologically similar PDC members , such as PhoQ , DcuS , CitA , DctB , and CzcS sensors , exhibit an enormous sequence variability ( S2 Fig ) . Within the common structural arrangement , the sequence varieties enable them to detect diverse stimuli . Most of the characterized PDC sensor domains ( CitA , DcuS , and DctB ) detect their cognate ligands by the internal cavity that is formed by the central conserved β-sheets ( Fig 1C ) [20 , 22] . Differently , the extracellular sensor domain of CzcS utilizes the residues from the symmetrical N-terminal α-helices to coordinate with Zn ( II ) in a tetrahedral coordination geometry . This constitutes a special class of Zn ( II ) binding sites that form at the dimer interface in the biochemical Zn ( II ) sites [24] . The recently reported metal-ion sensor CusS binds the effector at the dimer interface as well . However , the CusS SD interacts with the Ag ( I ) by using the N-terminal and C-terminal α-helices separately from different monomers . In the crystal structure of CzcS-Zn , the average distance is approximately 14 . 9 Å between the H1 and H1’ α-helices ( Fig 2A ) . In the absence of outer-shell constraints ( S9A Fig ) , the H1 and H1’ α-helices are flexible in rearranging the structural orientation . Their minor reorientation will initiate a large readjustment that affects the Zn ( II ) binding site ( S9B Fig ) . These structural features enable the rapid regulation of the active site for Zn ( II ) binding or releasing . It is similar to that of the Zn ( II ) binding site confined between TM2 and TM5 in the Zn ( II ) transporter YiiP [41] . In the CzcS SD , the coordination environment is symmetrical with the His55 , Asp60 , His55’ , and Asp60’ residues from the H1 and H1’ α-helices , respectively ( Fig 2A ) . This class of Zn ( II ) ligands that comprise the His and Asp residues is rare for the reported biochemical Zn ( II ) sites [42] . It makes the CzcS SD bind to Zn ( II ) with an affinity of 1 . 7 ( ±0 . 2 ) ×10−6 M . The RT-PCR analysis also indicates that the expression levels of czcS , czcR , czcC as well as oprD have obviously up-regulation or down-regulation when the P . aeruginosa is stimulated by Zn ( II ) at a micromole level ( S10 Fig ) . When the Zn ( II ) ligands of HK CzcS are substituted by cysteine residues , the mutants H55C and D60C can respond to Zn ( II ) stimulus as well and bind Zn ( II ) with higher affinities than that of the wild type construct in vitro . For the D60C mutant , a classic Cys2His2 zinc finger configuration with a tetrahedral coordination geometry ( S11A Fig ) can be properly formed by the residues Cys60 , Cys60’ , His55 , His55’ with Zn ( II ) [43] . The similar coordination geometry to that of the wild type HK CzcS causes the D60C mutant to display equivalent activities in sensing and regulating Zn ( II ) signal ( Fig 2C ) . We speculate that a linear coordination geometry may be formed by Cys55 and Cys55’ with Zn ( II ) in the H55C mutant , which is similar to the configuration formed on the dimer interface of the colicin E3 immunity protein ( S11B Fig ) [42 , 44] . The H55C mutant strain maintains the ability to respond Zn ( II ) stimulus as well ( Fig 2C ) . The biologically relevant dimer is observed in the crystal structure of CzcS-Zn complex with Zn ( II ) binding at the dimer interface ( Fig 1B ) . We ever made great efforts but failed in crystallizing the CzcS SD in the absence of Zn ( II ) . The difficulties in crystallization may be predominantly caused by the high flexibility of the CzcS SD especially the swing of N-terminal α-helices . In the absence of Zn ( II ) , the CzcS SD is exited as monomer in solution . When it binds to Zn ( II ) , the CzcS SD transforms from monomer to dimer and seems to be more conformational stable with the H1 and H1’ α-helices confined by Zn ( II ) coordination . Along with in vivo biological evidences , the Zn ( II ) induced dimerization of the CzcS SD is supposed to be physically important for signal regulation . This speculation is also confirmed by the different effects of Co ( II ) on the regulation of antibiotic resistance between wild type and mutant D60C strain ( Fig 3 ) . The D60C strain turns to be a Co ( II ) -responsive regulator that shows Co ( II ) -inducible resistance to MEPM . In vitro , the crosslinked dimerization of sensor domain induced by Co ( II ) is also observed for the D60C mutant . These experiments again indicate that the association of H1 and H1’ α-helices is necessary for the activity of HK CzcS . What’s more , the association state of H1 and H1’ α-helices maintains till to the linker region when HK CzcS is in the activated form . In the case that the original Zn ( II ) binding site is destroyed ( mutant H55A ) , a cysteine substitution in the linker region ( mutant L38C H55A ) can bind to Zn ( II ) with high affinity in vitro and strongly respond to the Zn ( II ) stimulus in vivo as well ( Fig 4 ) . The rigid structural features of H1 and H1’ α-helices are the pivotal guarantee for the signal transduction , and this is verified by the proline substitutional experiments . Proline residues are known to distort the structure of helices [45 , 46] . Visual inspection of some helices with a proline residue demonstrates a range of helix distortion ( S12 Fig ) . Obvious kink angle is observed in some proline-containing α-helices ( S12A Fig ) , which may happen to the H1 and H1’ α-helices in the mutant strains L50P , N45P , R43P , and R41P . This kind of distortion of H1 and H1’ α-helices makes the mutant strains ( L50P , N45P , R43P , and R41P ) seriously impair in the functions of Zn ( II ) induced metal detoxification and antibiotic resistance . While , there also exist some proline-containing α-helices that are approximately straight ( S12B Fig ) . It may be the reason that why the growing status of mutants Q52P and L48P is not influenced by the introduction of proline residue on H1 and H1’ α-helices . In addition , the H1 and H1’ α-helices can’t resist the double proline substitutions , which lead serious impairments in the abilities of responding to Zn ( II ) signal . Thus , it’s important to keep the conformation of H1 and H1’ α-helices in the signal transduction . As above described , the H1 an H1’ α-helices are the key factors in the activity of HK CzcS . They interact with the Zn ( II ) and keep the Zn ( II ) -induced association state till to the linker region , which are physiologically important for extracellular signal sensing and transduction across the transmembrane helices to the cytoplasmic kinase ( Fig 6 ) . By other group research , the rearranged helical interactions are discovered within the dimeric four-helical bundles in the transmembrane domain when HK CzcS is activated [47] . There is a transition from intramolecular- to intermolecular-crosslinking within the transmembrane helices ( Fig 6 ) [47] . We speculate it’s the association of H1 an H1’ α-helices that leads the structural rearrangements in the sensor domain , which will drive the interactional displacements of the helix bundles in the transmembrane domain . The aforementioned quaternary structural changes within the homodimer ultimately lead the trans autophosphorylation in cytoplasmic kinase domain on the conserved histidine residues ( Fig 6 ) [48–51] . The promising model ( Fig 6 ) presented here provides preliminary insights into the molecular mechanism of Zn ( II ) signal sensing and transduction by HK CzcS . It gives an implication for understanding the Zn ( II ) induced metal detoxification and antibiotic resistance in CzcR-CzcS TCS of P . aeruginosa . However , besides the structural information of extracellular sensor domain provided in this study , the structural characterizations of transmembrane domain and cytoplasmic kinase domain have not been reported to date for HK CzcS . Thus , further investigations are still needed to precisely characterize how the signal in the sensor domains results in the interactional rearrangement of the transmembrane helices and modulates the autophosphorylation events in the cytoplasmic kinase domains .
The DNA fragment of the CzcS SD ( amino acids 40–166 in CzcS protein ) was amplified from P . aeruginosa genomic DNA and cloned into the NheI and HindIII sites of the pET-28a vector ( named pCSET ) . E . coli BL21 ( DE3 ) cells transformed with the construct were cultivated in the LB medium supplemented with 30 μg/ml of kanamycin . The cells were cultivated at 37°C with constantly shaking at 250 rpm following a 1:100 inoculation from an overnight culture . Expression was induced with 0 . 5 mM IPTG when the culture reached an optical density of OD 600 ≈ 0 . 6 . The induced cells were grown for 4 h at 30°C , and subsequent steps were performed at 4°C . Cells expressing the CzcS SD with an N-terminal His6-Tag were harvested by centrifugation and lysed by sonication on ice in 15 ml of lysis buffer ( 10 mM Tris-HCl , pH 7 . 4 , 100 mM NaCl , 0 . 1 mM PMSF , 10% glycerol , and 1 μl DNaseI ) . The supernatant was obtained by centrifugation at 12 , 000 rpm for 15 min and loaded onto a 5-ml Ni-NTA column that was pre-equilibrated with 2–3 column volumes of buffer A ( 10 mM Tris-HCl , pH 7 . 4 , 100 mM NaCl , and 25 mM imidazole ) . The fusion protein was eluted in a linear gradient with the concentration of imidazole ranging from 75 mM to 500 mM in buffer A . The N-terminal His6-tag was removed by digesting the fusion protein with a protease overnight . The His6-tag-cleaved protein was treated with 0 . 5 mM EDTA and purified on an 8-ml MonoQ anion-exchange column which was equilibrated with buffer B ( 10 mM Tris-HCl , pH 7 . 4 , and 50 mM NaCl ) . The protein was eluted from the MomoQ column with 100 mM NaCl in buffer B . The purified CzcS SD were identified by SDS-PAGE and used in the following experiments . The CzcS SD mutants ( CzcS SD H55C and CzcS SD D60C ) were obtained by using site-directed mutagenesis technology performed on the pCSET plasmid . The CzcS SD L38C H55A mutant ( amino acids 27–175 in CzcS protein ) was constructed in the same method as pCSET plasmid followed by site-directed mutagenesis of sites Leu38 and His55 . All the CzcS SD mutants were expressed and purified in the same procedures as wild type CzcS SD . To crystallize the CzcS-Zn , the CzcS SD with the concentration of 3–4 mg/ml was mixed with an equimolar amount of ZnSO4 in the buffer containing 10 mM Tris-HCl , pH 8 . 5 , and 100 mM NaCl . The complex crystals were grown at 16°C by the sitting-drop vapor-diffusion against the reservoir buffer containing 0 . 2 M ( NH4 ) 2SO4 , 0 . 1 M Bis-Tris pH5 . 5 , and 25% w/v polyethylene glycol 3350 . The irregular cuboid crystals came out after two days and continued to grow until reaching a suitable size for X-ray diffraction studies . The crystals were briefly soaked in a cryoprotectant containing 0 . 2 M ( NH4 ) 2SO4 , 0 . 1 M Bis-Tris pH5 . 5 , 25% w/v polyethylene glycol 3350 , and 8% glycerol prior to flash-frozen in liquid nitrogen . The diffraction datasets at the zinc K-edge were collected from single crystals at BL17U beamlines at the Shanghai Synchrotron Radiation Facility [52] . The X-ray diffraction datasets were integrated and scaled with the HKL2000 package software . The initial phase for automated model building was solved by zinc single-wavelength anomalous dispersion using the Phenix software [53] . Iterative rounds of refinement were performed using the Phenix software , which were followed by manual alterations using the WinCoot software [54] . Refinement was conducted until no significant improvements were achieved . All structural models for the current study were generated with the PyMOL software [55] . The data collection and refinement statistics are listed in Table 1 . The atomic coordinates and structural factors for CzcS-Zn have been deposited into the Protein Data Bank with accession code 5GPO . The czcS-deficient strain of P . aeruginosa was constructed with a homologous recombination assay [56] . A 2025-bp PCR fragment corresponding to the first 8 bp of the czcS gene was amplified from P . aeruginosa genomic DNA with the primers I and II which contain an EcoRI and an XbaI restriction sites , respectively . Another 1892-bp PCR fragment that contains the 3’-end of the czcS gene was amplified from P . aeruginosa genomic DNA with primers III and IV which contain an XbaI and a HindIII restriction sites , respectively . The intervening gentamicin resistance cassette was amplified from the pPS858 plasmid with the XbaI restriction site both at 5’- and 3’-end . The aforementioned three DNA fragments were ligated into an EcoRI /HindIII-cleaved pEX18AP plasmid . The constructed plasmid was transformed into the P . aeruginosa competent cells by electroporation [56] . The successfully homologous recombinants were screened on the LB medium containing 30 ug/ml gentamycin . The czcS-deficient strain was further identified by PCR and DNA sequencing . The oligonucleotides used to construct the czcS-deficient strain are listed in S3 Table . The czcS operon and its encoding gene were amplified from P . aeruginosa genomic DNA . They were ligated by overlap PCR and cloned into the HindIII/BamHI restriction sites of pAK1900 . The complementary plasmid pCSAK was verified by DNA sequencing and used to construct the mutants . Site-directed mutagenesis of CzcS was performed on the yielding pCSAK plasmid using the Quik Change site-directed mutagenesis kit ( Agilent Technologies ) . All the variants were verified by DNA sequencing . The oligonucleotides used for the site-directed mutagenesis plasmids are listed in S3 Table . The pCSAK plasmid was transformed into the czcS-deficient strain by chemical transformation to supply as the complementary strain . The empty pAK1900 plasmid was transformed into the wild type P . aeruginosa and czcS-deficient strain to supply as the positive control and negative control , respectively . All the variants were separately transformed into the czcS-deficient strain . The aforementioned strains were cultivated in LB mediums supplemented with 150 μg/ml carbenicillin . They were grown overnight at 37°C with constantly shaking at 250 rpm/min . Fresh LB mediums containing 150 μg/ml carbenicillin were inoculated with the overnight cultures at the proportion 1:100 and grown at 37°C until the density of OD 600 arrived 1 . 0 . The cultures then underwent ten-fold serial dilutions with five gradients and were further seeded onto the LB plates with varying concentrations of Zn ( II ) , Co ( II ) or MEPM . The plates were incubated at 37°C for 16 h before observation . The chromogenic chelating agent PAR was selected as the competitor in the spectrometric determinations of binding affinity of Zn ( II ) with wild type and mutant CzcS SD ( CzcS SD H55C , CzcS SD D60C , and CzcS SD L38C H55A ) . All Zn ( II ) binding experiments were performed under photophobic condition at 22°C in the buffer containing 10 mM Tris-HCl , pH 7 . 4 , 150 mM NaCl , and 0 . 5 mM TCEP ( for the mutants CzcS SD H55C , CzcS SD D60C , and CzcS SD L38C H55A only ) . A known concentration of PAR solution ( 36 uM ) was mixed with the purified protein ( 50 uM-200 uM ) . The mixtures were divided into equal volumes followed by loading consistent volumes of ZnCl2 with increasing concentrations ( in the range of 2 uM-42 uM ) . The UV-visible spectrum were recorded in the range of 200 nm to 700 nm until the reaction systems achieved competitive equilibrium . The titration data at 500 nm were fit with Dynafit software [30] by using one-site model to obtain the apparent dissociation constants of Zn ( II ) with wild type and mutant CzcS SD . The CzcS SD was purified in the same procedures as described above in the buffer containing 10 mM HEPES , pH7 . 4 , 100 mM NaCl . The primary amine reactive crosslinker BS3 was stored in DMSO at 100 mM and diluted to 1 mM in 20mM HEPES ( pH7 . 4 ) immediately before use . A 50-fold molar excess of BS3 crosslinker was loaded into the CzcS SD and CzcS-Zn samples with a final concentration of 1 mM . The reaction systems were incubated at room temperature for 30 minutes and quenched with 50 mM Tris-HCl , pH 7 . 4 . The quenching reaction was incubated at room temperature for 15 minutes . The CzcS SD , CzcS-Zn , and the products of the crosslinking reactions were analyzed by 14% SDS-PAGE and quantified by ImageJ [34] . The crosslinking experiments of other divalent cations , such as Mg ( II ) , Mn ( II ) , or Co ( II ) , were performed in the same procedure . The quantitative real-time RT-PCR with the rpsl gene as the reference was performed to monitor the expression changes of czcS , czcR , czcC and oprD genes when P . aeruginosa is stimulated by the Zn ( II ) at a micromolar level . The total RNA was extracted by traditional phenol-chloroform method and reverse transcribed by iScript cDNA Synthesis Kit ( Bio-Rad ) . The cDNA samples were diluted for different folds and used as the templates in the PCR experiments . The real-time RT-PCR was performed on the Bio-Rad CFX96 equipment using the Ssofast SYBR Green Supermix ( Bio-Rad ) . The experiments were performed at least three independent times with average results shown . The primer sequences used for real-time RT-PCR are designed using the Primer3 program and listed in S3 Table . | P . aeruginosa inhabits diverse environments and is one of the most prevalent opportunistic human pathogens of immunocompromised patients . The high antibiotic resistance is a major cause of therapeutic failure in the treatment of P . aeruginosa infections . The opportunistic pathogen P . aeruginosa co-regulates cross-resistance between Zn ( II ) and carbapenem antibiotics by the CzcR-CzcS signal transduction system . The extracellular Zn ( II ) stimulus is sensed by the HK CzcS and further triggers metal detoxification and antibiotic resistance through intracellular regulatory pathway . Here , we provide the three-dimensional structure of CzcS SD in complex with the Zn ( II ) . Based on the structure , several key residues for Zn ( II ) sensing and regulation are identified , and the signal transduction is disclosed to be modulated by the dimerization of N-terminal α-helices in the sensor domain . Our research will provide potential guidance for the treatment of clinical issues caused by co-regulation between heavy metals and antibiotics in P . aeruginosa . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"antimicrobials",
"dimers",
"(chemical",
"physics)",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"pathogens",
"drugs",
"microbiology",
"organic",
"compounds",
"toxicology",
"pseudomonas",
"aeruginosa",
"a... | 2017 | Structural basis of Zn(II) induced metal detoxification and antibiotic resistance by histidine kinase CzcS in Pseudomonas aeruginosa |
In motor tasks with redundancy neuromotor noise can lead to variations in execution while achieving relative invariance in the result . The present study examined whether humans find solutions that are tolerant to intrinsic noise . Using a throwing task in a virtual set-up where an infinite set of angle and velocity combinations at ball release yield throwing accuracy , our computational approach permitted quantitative predictions about solution strategies that are tolerant to noise . Based on a mathematical model of the task expected results were computed and provided predictions about error-tolerant strategies ( Hypothesis 1 ) . As strategies can take on a large range of velocities , a second hypothesis was that subjects select strategies that minimize velocity at release to avoid costs associated with signal- or velocity-dependent noise or higher energy demands ( Hypothesis 2 ) . Two experiments with different target constellations tested these two hypotheses . Results of Experiment 1 showed that subjects chose solutions with high error-tolerance , although these solutions also had relatively low velocity . These two benefits seemed to outweigh that for many subjects these solutions were close to a high-penalty area , i . e . they were risky . Experiment 2 dissociated the two hypotheses . Results showed that individuals were consistent with Hypothesis 1 although their solutions were distributed over a range of velocities . Additional analyses revealed that a velocity-dependent increase in variability was absent , probably due to the presence of a solution manifold that channeled variability in a task-specific manner . Hence , the general acceptance of signal-dependent noise may need some qualification . These findings have significance for the fundamental understanding of how the central nervous system deals with its inherent neuromotor noise .
Decrease of error and its variability as a consequence of practice is a widely recognized indicator of skilled performance and improvement . More recent studies have tried to look beyond pure outcome measures and examined the variability at different stages in movement generation , for example during the planning stage [1] , during the execution of movements [2] , [3] , and in the processing of sensory estimates [4] . Such variability or noise is the consequence of many processes at all spatiotemporal levels of the sensorimotor system arising , for example , in signal propagation due to synaptic fluctuations that affect the regularity of spike trains , or in the transduction of a continuous signal into discrete spike sequences [5] . This variability has been shown to depend on the signal amplitude , for example the magnitude of contractile force or velocity . It has become widely accepted that subjects aim to minimize signal-dependent noise [6] , [7] . Over recent years sensorimotor noise and its role in motor control has received increasing attention from several lines of study . For example Trommershäuser , Maloney and colleagues have focused on rapid pointing tasks where variability in pointing accuracy was analyzed with respect to different penalties and rewards [8] , [9] . Several studies have shown that human performers take their variability and the risk induced by their own uncontrolled variability into account . Their research has been guided by the framework of decision theory and emphasized the cognitive decision making and planning when performing a motor task . Van Beers and colleagues have looked at variability of reaching tasks as an entry to understand visual and proprioceptive information contributing to motor solutions [3] , [10] . Variability and noise is also central in the work on stochastic optimal feedback control by Todorov and colleagues and this computational approach has been applied to increasingly more diverse tasks [11] , [12] , [13] , [14] . A recent study by Nagengast , Braun , and Wolpert highlighted that this optimal control framework may need to be differentiated to address inter-individual differences in risk attitudes , i . e . , individuals' preferences to deal with risk and penalties [15] . Our research on variability and noise complements and extends these lines of research in several aspects . The present study examines performance of a motor skill where redundancy in the task presents different opportunities for dexterous performance . To be explicit , redundancy in the task permits that an infinite set of executions leads to the same result , both for zero-error solutions but also all other non-zero task solutions . This redundancy has been frequently illustrated in a multi-joint pointing movement where an infinite number of joint-angle combinations lead to a given accuracy in the endpoint position . In our single-joint throwing task an infinite set of states at the moment of ball release , position and velocity of the arm movement , leads to zero-error performance . However , not all solutions are the same with respect to risk and sensitivity to error . Mathematical analysis of the task's redundancy presents the platform for an analysis of subjects' variability over repeated executions . Repetitions of the “same” movement will lead to variations not only as a consequence of the ever-present noise in the sensorimotor system but also due to the geometry of the null space of the task that endows different solutions with different degrees of tolerance or sensitivity to errors . Hence , the observed variability is not necessarily random , but rather its distribution may express strategies of the central nervous system . Our analysis will focus on distributional aspects of execution with respect to the geometry of the null space or solution manifold determined by the task . Related approaches such as decision-theoretic , optimal control models , or the UnControlled Manifold ( UCM ) method have provided support that the variability over multiple repetitions is structured . For example the UnControlled Manifold ( UCM- ) approach [16] , frequently applied to variability in joint space with respect to its mean endpoint position or force contributions of fingers with respect to summed force output has provided support that variability in directions parallel to the null space is larger than variability orthogonal to it [17] . Interestingly , this structured variability is also the consequence of the optimization of control cost in the optimal feedback control models [13] . While the goal of the UCM-analysis resembles our approach , some critical differences exist in how the problem is posed , how variability is analyzed and , consequently , the obtained result [18] , [19] . The present study illustrates our approach and how it permits specific predictions about strategies with a view to a desired task result . One critical difference between our approach and the UCM-method and optimal feedback control is that they have only focused on the covariance structure of the distribution with respect to a solution manifold . In contrast , our work developed an analysis of variability that differentiates between three different contributions to optimal task performance . This TNC-method allows the quantitative analysis of Tolerance , Noise and Covariation [19] , [20] , [21] , [22] . The component Noise is straightforward and refers to the amplitude of the random distribution . Covariation is indicated when the data are aligned with the solution manifold , conceptually identical to what the UCM method and also optimal feedback control focused on . Our quantification , however , does not rely on the analysis of the covariance structure which is stricken with sensitivity to coordinates [18] . Unique to our analysis is the concept of Tolerance that evaluates movement strategies with respect to the error that deviations from the ideal solution incur , i . e . , tolerant solutions are least sensitive to error and perturbations . It should , however , be pointed out that this concept is not equivalent to local sensitivity as Tolerance is defined over the neighborhood defined by the subject's variability . Note also that maximizing Tolerance is different from the goal of “maximizing hit rate” in a single trial by processing feedback to decrease error . Rather , it is defined over a set of performances and quantifies to what degree subjects are sensitive to their own errors and take predicted cost of a set of trials into account . Previous experiments have shown how Tolerance is the first component that is reduced with practice [20] , [22] . The present study shows how a task analysis can generate predictions that permit direct evaluation of whether subjects seek out error-tolerant strategies , i . e . , strategies that allow maximum variability at the execution level but with minimal penalty in the result . To this end we examine a throwing task called skittles in which a subject throws a ball suspended to a vertical post to hit a target skittle at the other side of the post . The task is redundant such that an infinite set of variations can have the same result . In the experimentally controlled task two execution variables , angular position and velocity at release of the ball , fully determine its result variable , the ball's trajectory and its error from hitting the skittle . The key characteristic is that the number of execution variables is larger than the number of result variables; hence , an infinite number of angle and velocity combinations can lead to the same distance error . The results with zero error form a set called the solution manifold . Hence , this task is representative for any goal-oriented skill where a redundant number of execution variables fully determine the result . This study examined the hypothesis whether subjects are sensitive to their motor variability and find error-tolerant solutions that minimize the effect of this variability on their performance result ( Hypothesis 1 ) . Yet , in the present task successful throwing actions can be executed with a large range of different velocities . As it is commonly assumed that higher velocities are associated with higher costs , such as signal-dependent noise or some form of energy or effort , it can also be hypothesized that subjects seek solutions with the lowest possible velocity ( Hypothesis 2 ) . Two experiments with different task configurations will test these two hypotheses .
Prior to data collection , subjects were instructed about the experimental procedure upon which they signed an informed consent form in agreement with the Institutional Review Board of the Pennsylvania State University . The task for the present study is similar to the game skittles or tetherball where the person throws a ball suspended as a pendulum around a pole to hit a target at the opposite side of the post . The trajectory of the ball is fully determined by the angular position and velocity at release of the ball and the mathematical relationship is modeled using basic mechanics [22] . After release , the ball trajectory describes an elliptic trajectory around a center post from which it is suspended ( centripetal force field ) . Performance results or errors are quantified by the minimal distance between the ball trajectory and the target skittle . Figures 1A and 1B shows a top down view as subjects saw it during the two experiments , respectively . A successful hit with zero error meant that the center of the ball went through the center of the target . In case they did not hit the skittle , the error was calculated as the minimum distance between the trajectory and the center of the target . In both examples two of the three trajectories illustrate how different combinations of the two execution variables can lead to the same result ( error = 0 ) , i . e . , the task is redundant . The redundancy relation between execution and result variables is captured in the execution space ( Figure 1C and 1D ) where every throw , defined by the variables angle and velocity , corresponds to one point . Different levels of success , quantified in the error , are displayed by different grey shades . Perfect hits with zero error are displayed in white and form the one-dimensional solution manifold; solutions with increasing error are shown by increasingly darker grey shades; black denotes a post hit . As the two constellations exemplify , different positions of the target and the center post create very different execution spaces and solution manifolds . Participants stood in front of a back projection screen operating a lever arm that simulated the throw ( Figure 2 ) . The height of the lever was adjustable for each person so that his/her forearm was placed horizontally with the elbow joint aligned with the axis of rotation . At the distal end of the manipulandum , the participant grasped a ball and closed a contact switch with his/her index finger . Extending the index finger corresponded to opening the grasp to throw the ball; this opened the switch and triggered the release of the ball on the visual display . The rotation of the manipulandum was measured by a potentiometer ( Vishay Spectrol , CA ) with a sampling rate of 650 Hz . The participant could stand to the right or left of the vertical fixation , throwing in clockwise or counterclockwise direction , depending on the task . The visual display ( 60 Hz update rate ) was presented on a back projection screen ( 1 . 80 m×1 . 40 m ) positioned 0 . 60 m in front of the participant . On the screen he/she saw the virtual lever arm moving in real-time that threw a ball to hit a target skittle on the other side of the center post . The ball trajectory was computed from the online measurements of angular position and the numerically differentiated velocity at release according to the model described in [22] . The ball's trajectory was displayed for 1 s after release , which was sufficient to provide visual feedback about the success of the throw . The data acquisition and the visual display were programmed in Visual C++ ( Microsoft , v6 . 0 ) ; the virtual display was implemented by Open GL Graphics ( Silicon Graphics , v1 . 2 ) . Participants were instructed to hit the target with the virtual ball as accurately as possible . After a self-timed short break subjects initiated the next trial . Typically , one trial including the break between trials lasted approximately 6 s . The throwing movement itself lasted approximately 350 ms . In Experiment 1 subjects performed three sessions , each consisting of three blocks with 60 trials in one block , yielding a total of 540 trials; in Experiment 2 each subject performed five sessions , giving a total of 900 trials . Between each block , participants rested for a few minutes . The total duration of each session was approximately 15 min . The sessions were collected on three and five consecutive days , respectively . In session 1 participants were instructed to try different release angles and release velocities to find successful strategies that achieved reliable solutions . In the subsequent sessions participants were instructed to no longer explore but to continue with the strategy that had proven most successful . Note , by strategy we do not mean that subjects necessarily have to repeat a single solution , but rather stay in the ballpark of solutions . They were encouraged to fine-tune their performance and avoid hitting the center post as they would receive a large penalty . A third control experiment was performed to examine whether performance of the throwing action without a target resulted in different levels of variability that depended on the release angle or velocity . In this Experiment 3 six subjects were asked to perform the same throwing movements , only that there was no target skittle . The instruction to the subjects was to perform the throwing movements at their preferred velocity but also at two higher velocities and two lower velocities than preferred . The only constraint was to avoid hitting the center post . Subjects performed five blocks of 25 trials each , each block with one of the five instructed velocities . The sequence of blocks was randomized across subjects . In Experiment 1 subjects saw the workspace as shown in Figure 1A with the target located at coordinates ( 35 , 125 cm ) and the center post with a radius of 33 cm located off center at ( 10 . 5 , −60 cm ) . The target skittle ( radius 1 . 50 cm ) was located at ( 35 , 125 cm ) . The ball radius was 2 . 50 cm . Figure 1C represents the associated execution space with the nonlinear solution manifold ( error = 0 cm ) , shown in white . Although each solution on the manifold is equivalent , different locations on the solution manifold have very different sensitivity or tolerance to errors , as illustrated by the changing curvature of the result function adjacent to the solution manifold . For reference , if the trajectory only touched the target , the error was 4 cm . If the ball hit the center post , the trial was penalized with the relatively large error of 60 cm , shown in black . The execution space for Experiment 1 was so designed that successful solutions could take on a relatively large range of release angles and the curvature at the solution manifold showed a pronounced change: smaller release angles showed higher tolerance to error – the curvature of the result function was shallow . Additionally , the most tolerant region transitioned discontinuously to one associated with large penalty – strategies that resulted in post hits . Solutions that allowed for a relatively large dispersion were adjacent to solutions that are penalized heavily – risky strategies . In Experiment 2 , subjects saw the workspace as shown in Figure 1B . The target was located at the coordinates ( 5 . 0 , 105 . 8 cm ) and the post was slightly smaller ( radius 25 cm ) but centered at the origin ( 0 , 0 ) . Figure 1D represents the associated execution space with the solution manifold that was approximately parallel to the velocity dimension , i . e . execution strategies were only little sensitive to velocity . This sensitivity or tolerance to variations in angle increased for higher velocities , although the gradient was relatively small . Importantly also , the solution with the lowest velocity was adjacent to the penalized post hits and therefore posed a risky strategy . To test the two hypotheses we performed simulations to render quantitative predictions for tolerant solutions . For Hypothesis 1 the error-tolerance T of all possible executions , i . e . angle-velocity ( α , v ) pairs , was computed . As Tolerance T is defined for a given distribution of data , we used the average standard deviations of all subjects in the present two experiments , determined a posteriori from all participants as a representative distribution . While an estimate of variability based on previous experiments would have served this purpose , a more accurate estimate was obtained from the actual standard deviations . In Experiment 1 these standard deviations were SDα = 11 . 70 deg and SDv = 40 . 49 deg/s , as determined from the grand average over sessions 2 and 3 . In Experiment 2 , these standard deviations were SDα = 9 . 44 deg and SDv = 70 . 38 deg/s , calculated over sessions 2 to 5 . These dispersions defined the size of the neighborhood for each location in execution space where i = 1 , 2 , …360 denotes one bin in the angle dimension , and j = 1 , 2 , … 360 denotes one bin in the velocity dimension ( see Text S1 for more detail ) . For each Tolerance was calculated as a weighted average error , . The weights over this matrix neighborhood were taken from a bivariate Gaussian distribution . was assigned the weighted average ( see Text S1 for details ) . To translate these Tolerance values into an estimate of probability by which subjects chose this strategy , was transformed by an exponential function , the softmax activation function , to obtain the expected results E ( R ) for each result [23]: The denominator is a normalization factor that scaled the values of E ( R ) to the range [0 , 1] . The parameter a was fitted based on the pooled data distributions using least square fits . This transformation paid tribute to the fact that the subjects' probability of choosing a given strategy did not scale linearly with the expected Tolerance . Rather , solutions with small error were given high preference , while solutions with intermediate and large errors were much less preferred and thereby less probable ( see Text S1 for details ) . For Hypothesis 2 – predicting preference for the velocity-sensitive strategy – the initial Tolerance estimates for each were also transformed by the softmax activation function . However , this transformation included an additive term that evaluated velocity v: Analogous to Hypothesis 1 , the two parameters a and b were fitted to the pooled data distributions using least square fits ( see Text S1 for details ) . Figure 3 illustrates the data distributions in both experiments and the two quantitative predictions for both experimental target constellations . The top two panels show the histograms of all subjects' data pooled , plotted on the respective execution space ( compare to Figure 1C and D ) . These histograms provided the reference for parameterizing the softmax function for the quantitative predictions . The two middle panels show the predictions of Hypothesis 1 . For Experiment 1 the maximum value of E ( R ) with highest Tolerance was at α = −44 deg and v = 161 deg/s ( indicated by the red circle ) . For Experiment 2 the different target constellation rendered the maximum of E ( R ) and highest Tolerance at the highest velocity for the given range: α = −82 deg and v = 1000 deg/s . It should be pointed out that the slope was very gradual and for higher velocities the change in E ( R ) was very small . Note that the exponential transformation decreased E ( R ) for intermediate or lower result values , thereby enhanced the contrast between good and less good solutions . The two bottom panels of Figure 3 show the simulation results for Hypothesis 2: For Experiment 1 the predicted optimal strategy was at α = −29 deg and v = 122 deg/s . While this optimum was close to the one of Hypothesis 1 , the gradient around it was much steeper . For Experiment 2 , the strategy with minimum velocity was at α = 83 deg and v = 142 deg/s . In this experiment , the two hypothesized solutions were at opposite ends of the manifold . To evaluate the subjects' distributions several analyses were conducted . First , to visualize each individual's distribution in execution space the covariance matrix of the execution variables was calculated and shown by its 95% confidence ellipse . Three parameters described the confidence ellipse: 1 ) the mean of release angle and velocity determined the center of the ellipse , 2 ) the eigenvectors were calculated to determine the orientation of the ellipse , and 3 ) the square roots of the eigenvalues determined the size of the semi-major and semi-minor axes of the ellipse . Given that the confidence ellipse required a large number of samples the data of all sessions , except session 1 were pooled . To test the two hypotheses , a first simple test evaluated how many confidence ellipses , i . e . , subjects , overlapped with the predicted optimal value of Hypotheses 1 and 2 . This resulted in a simple count that was compared with an expected frequency derived under the assumption that there was no preference for any specific solution . A second more thorough test examined each individual's distribution and compared it with the hypothesized distribution at the respective location in execution space . To this end , the trial distributions of each subject ( 360 trials in Experiment 1 and 720 trials in Experiment 2 ) were presented in execution space in a matrix of 5x5 cells centered on the mean angle and velocity; the matrix size was determined by the individual's standard deviations . The number of cells for the matrix was based on the recommended √n , which suggested 18 cells for Experiment 1 and 27 cells for Experiment 2 . To facilitate comparison of results for Experiments 1 and 2 we chose 25 cells , or a 5x5 matrix for both . The frequency distributions of the data were compared with the predictions for E ( R ) from Hypotheses 1 and 2 using likelihood estimates . Given that the predictions for Hypothesis 2 contained two fitting parameters , it was evident that Hypothesis 2 had to fare better . Hence , for the comparison of the two nested model fits , we applied the Akaike Information Criterion AIC that evaluated the goodness of fit in the face of different parameters .
These reviewed studies on sources of variability discussed the presence or absence of variability in terms of its amplitude and generally implied a random structure . While it is beneficial if this noise amplitude is reduced , the nervous system has also found other ways to reduce undesired variability in the behavioral outcome . If a given task is redundant , one such way is to channel variability into directions that have little effect on the end result . For example , the linkage of joints in the arm may covary without necessarily affecting the outcome , as shown for example in pistol shooting [25] , [26] , [27] , [28] , dart throwing , Boule throwing [29] and basketball throwing [30] . Much experimental evidence has been accumulated for this phenomenon and covariance has been generalized as a signature of synergies [31] . Channeling of variability into “do-not-care” directions is also an important consequence of stochastic optimal feedback control as applied to motor tasks by Todorov and colleagues [32] , [33] . Our previous studies have similarly shown covariation in the structure of variability , although our three-pronged approach differentiated between magnitude and anisotropy of the data distribution ( Noise and Covariation ) . It also separated off Tolerance , the aspect that figured centrally in the present study [20] . Core to our task-based analysis is the distinction between execution and result space: by mapping executions into results , the layout and geometry of the results can be obtained . This not only offers a quantitative understanding of the zero-error solutions but also of their neighborhood and the curvature , i . e . their sensitivity to error . In previous work we introduced Tolerance , a concept that allows quantification of what is the optimal strategy for a given distribution or variability [19] , [22] . The present study extended this work by developing a priori predictions where and how variability should be distributed if the nervous system chose error-tolerant solution strategies . The first hypothesis was that in skilled performance actors are aware of the limited resolution in their control and take their variability into account when planning and executing a movement . This hypothesis was tested by calculating the expected result in a neighborhood around each solution , i . e . by quantifying the degree of tolerance of a given movement strategy . This approach differs from standard sensitivity analyses in linearized systems that assess the effects of small deviations from a single solution . Specifically , local linear stability analysis evaluates how ( infinitesimally ) small perturbations destabilize a solution; relaxation time provides a quantitative measure for how fast a system returns to the stable solution . However , such an approach is ignorant to the effects of slightly larger errors . Knowledge of an extended neighborhood , however , is important when the system is nonlinear and has discontinuities like the result space in the skittles task . Considering that in human performance perturbations or errors have a sizable variance and the result space is nonlinear as in our skittles task , it is appropriate to assess error sensitivity not only at a point , but in a neighborhood around a chosen solution ( for discussion of such analyses in nonlinear systems see [34] ) . The present study presented an analysis that quantifies error-tolerant strategies by assessing an “area” of solutions determined from the actual variability of subjects and evaluated the expected performance for such variability . Results of two different task variations supported that subjects seek error-tolerant strategies . In Experiment 1 the data distributions of all nine participants were best fitted by the predictions of Hypothesis 1 . However , the results did not rule out that subjects also minimized velocity as the solutions with the highest Tolerance were close to solutions at relatively low velocities . Interestingly , some individuals' strategies were also close to solutions with high penalty for hitting the center post . These inter-individual strategies may reflect the individual's attitude to risk , a topic that has been investigated by [35] . The rationale and the results of our study are in overall accordance with a series of experiments by Trommershäuser , Maloney and colleagues [9] , [36] , [37] . Using a speed-accuracy pointing task where the target area was bounded by a penalty area ( at different distances and with different penalties ) , the distribution of hits was examined with respect to the expected gain . Formalized in a decision theoretical framework where a gain function is optimized based on the weighted sum of the gain and the subject's inherent variability , the results showed systematic effects of the penalty on the distributions . The results therefore supported the conclusion that selection of a movement strategy is largely determined by the subject's inherent variability . In contrast to the present study , hitting success was binary ( positive for the target area and negative for the penalty area ) while our focus was on the continuous distance to the target , which was prerequisite to the sensitivity analysis central to our study . Importantly , in these experiments the reward or penalty was endpoint accuracy that was directly visible to the subject on a monitor . In the skittles task the variables at release were not visible and important variables needed to be learnt via proprioceptive information across repeated trials , not itself visible to the actor . Further , the solution manifold and the sensitivity of its neighborhood are highly nonlinear and it is unlikely that performers have a priori an internal model of the result space . Hypothesis 2 was formulated based on the widely accepted findings that performance variability scales with movement speed such that performance at higher velocities is more variable [38] , [39] , [40] . Assuming movement velocity reflects the amplitude of the motor control signal , this observation can be generalized that variability increases with signal strength and velocity . Physiologically , this behavioral observation has been related to the organizational properties of the motor unit pool such as recruitment order and twitch amplitudes [24] . In addition , it has also been commonly argued that subjects aim to minimize energy , either mechanical or metabolic . In the case of skittles , it may be hypothesized that subjects seek throws with minimum momentum of the arm movement . Taking these arguments together the hypothesis can be formulated that subjects should seek solutions with the lowest possible velocities [6] , [41] . This alternative hypothesis was tested in Experiment 2 that was designed to explicitly dissociate between the two hypotheses . The target configuration was modified to create an execution space that permitted a large range of velocities to achieve successful hits where Hypothesis 2 clearly predicted the lowest possible velocity as the preferred strategy . In contrast , error tolerance showed a maximum at the highest velocities , although the gradient of E ( R ) across the higher velocities was relatively small . The individual subjects' distributions did not provide support for Hypothesis 2 and the subject averages and confidence ellipses extended over a large range of velocities . Consistent with this finding , analysis of velocity-dependent variability revealed that across subjects the variability did not increase with mean velocity . To further scrutinize the apparent absence of velocity-dependent scaling of variability , an additional experiment was conducted to test whether this finding was due to the goal-oriented nature or the redundancy of the task . We speculated that if motor solutions cluster along the solution manifold this may obscure the otherwise reported increase in variability with movement velocity . In Experiment 3 subjects executed the same movement but did not aim for a target skittle . Hence , there was no solution manifold constraining the actions . This result highlighted that task redundancy introduces a solution manifold that presents a constraint that may suppress the velocity-dependent variability . This result is important as it qualifies the frequently adopted general assumption that variability and noise increases with signal amplitude . As a final comment , it should be pointed out that our approach is completely confined to the kinematic level of task performance . Limb dynamics or other biomechanical considerations are not taken into explicit consideration . This is justified on two counts: First , the skittles task is performed by only a single-joint rotation in the horizontal plane where the rotating joint is fixed to the axle of the lever arm . Hence , neither intersegmental torques in the executing arm nor gravitational influences are of immediate concern . Second , much research on upper limb movements has provided evidence that endpoint trajectories may be planned in kinematic extrinsic coordinates [42] , [43] . The analysis uses angular rotations of the manipulandum as defined in extrinsic coordinates with respect to the screen . That said , we did not address potential biomechanical considerations that arise from the positioning of the body with respect to the manipulandum or with what joint angles the angular rotations were executed . Subjects were told to position themselves in the most comfortable position , both with respect to any biomechanical concerns and with respect to optimal vision of the screen . At present , we refrained from including such additional considerations as these would have required additional motion capture . In summary , two experiments examined a virtual throwing task and presented an analysis that provided an a priori hypothesis about which strategies actors should employ if they optimized error-tolerance . Analysis of the relation between the variability in execution to the result of the task performance revealed that actors not only decreased their motor variability in execution variables that mattered for the success of the task . The findings also gave strong support that subjects were sensitive to their motor variability and preferred strategies that optimized error tolerance . | It is widely recognized that variability or noise is present at all levels of the sensorimotor system . How the central nervous system generates functional behavior with a sufficient degree of accuracy in the face of this noise remains an open question . This is specifically relevant when the motor task is redundant , i . e . , where many different executions can achieve the same task goal . Using an experimentally controlled throwing movement as model task we examined how humans acquire movement strategies that are tolerant to intrinsic noise . Based on a new computational approach that parses variability based on an analysis of task redundancy , we tested two hypotheses: 1 ) Subjects are sensitive to noise and seek solutions that are tolerant to this noise . 2 ) Subjects avoid solutions with high velocities and the costs associated with high velocities . Analysis of the distributional properties of variability in two experiments revealed that humans select those strategies that maximize error-tolerance . These findings have significance for fundamental understanding of the central nervous system and for learning in the context of rehabilitation . | [
"Abstract",
"Introduction",
"Methods",
"Discussion"
] | [
"medicine",
"social",
"and",
"behavioral",
"sciences",
"biology",
"engineering"
] | 2011 | Neuromotor Noise, Error Tolerance and Velocity-Dependent Costs in Skilled Performance |
Leptospirosis is an important re-emerging infectious disease that affects humans worldwide . Infection occurs from indirect environment-mediated exposure to pathogenic leptospires through contaminated watered environments . The ability of pathogenic leptospires to persist in the aqueous environment is a key factor in transmission to new hosts . Hence , an effort was made to detect pathogenic leptospires in complex environmental samples , to genotype positive samples and to assess leptospiral viability over time . We focused our study on human leptospirosis cases infected with the New Caledonian Leptospira interrogans serovar Pyrogenes . Epidemiologically related to freshwater contaminations , this strain is responsible for ca . 25% of human cases in New Caledonia . We screened soil and water samples retrieved from suspected environmental infection sites for the pathogen-specific leptospiral gene lipL-32 . Soil samples from all suspected infection sites tested showed detectable levels of pathogenic leptospiral DNA . More importantly , we demonstrated by viability qPCR that those pathogenic leptospires were viable and persisted in infection sites for several weeks after the index contamination event . Further , molecular phylogenetic analyses of the leptospiral lfb-1 gene successfully linked the identity of environmental Leptospira to the corresponding human-infecting strain . Altogether , this study illustrates the potential of quantitative viability-PCR assay for the rapid detection of viable leptospires in environmental samples , which might open avenues to strategies aimed at assessing environmental risk .
Leptospirosis is an acute febrile disease caused by pathogenic spirochetes of the genus Leptospira . It is considered an important re-emerging infectious disease that affects more than 1 million humans worldwide [1] . The spectrum of human disease caused by leptospires is extremely wide , ranging from subclinical infection to a severe syndrome of multiorgan infection with high mortality . Leptospira transmission from the urine of reservoir hosts to incidental hosts , including humans , usually occurs through the contamination of skin lesions or mucosae with contaminated surface water or soil [2] . The incidence of such infections depends on several factors including the density of the reservoir species and its Leptospira carriage prevalence , the dilution into watered environment and the survival time of the leptospires into possibly nutrient-poor and adverse environmental conditions . Estimation of survival time and virulence preservation of pathogenic Leptospira spp . after excretion into the environment is becoming a crucial challenge to determine the environmental risk and to adopt preventive measures . The duration of Leptospira survival in natural habitat is affected by many factors including abiotic and biotic factors . The persistence of pathogenic Leptospira in moist soil and freshwater for long periods of time is thought to depend on a slightly alkaline pH , high oxygen , and low salt concentrations [3–5] . The classical assumption is that slightly higher alkalinity ( up to pH 8 . 0 ) allows for longer survival . Under laboratory conditions , a strain of serovar Javanica was reported to survive in distilled water ( pH 7 . 8 ) for 152 days [6] . More recently , Andre-Fontaine et al . [7] showed that pathogenic Leptospira can survive for months in mineral water . Interestingly , Leptospira were reported to survive as long as 10 months in adverse conditions ( 4°C ) and up to 20 months when stored at 30°C . Interactions of Leptospira spp . with the environmental microbiota also begin to be examined . Environmental microbial blooms alter the concentration of oxygen , minerals , and other nutrients in the water and favor either multiplication or destruction of some species of pathogenic Leptospira [8] . Several common bacterial genera including Azospirillum and Sphingomonas were found along with pathogenic and saprophytic Leptospira spp . in biofilms formed in freshwater or in dental water unit systems [9 , 10] . Co-incubation with a Sphingomonas spp . increased Leptospira growth rate [8] , suggesting possible syntrophic interactions . When incubated with Azospirillum brasilense , viability of pathogenic Leptospira was enhanced at high temperature and extended under UV radiation or exposure to penicillin G , tetracycline or ampicillin . In addition , soil adsorption , thought to be an important step that favors leptospire persistence in the environment , was greatly increased in the presence of A . brasilense [8] . A major impediment to assess environmental risk for leptospirosis has been the difficulty to isolate pathogenic Leptospira from environmental samples , attributable in part to the fact that non-pathogenic leptospires outgrow pathogenic strains in culture . Other methods including direct animal inoculation are time-consuming , ethically questionable and have a low analytical sensitivity . However , the increasing use of molecular methods overcomes some limitations inherent to culture- and animal-based methods and provides quantitative information about the concentration of leptospires in contaminated waters [11–13] . New Caledonia provides an ideal location for studying environmental risk factors of leptospirosis because of its high leptospirosis incidence , on average 45 cases per 100 , 000 inhabitants , and the presence of known hot spots where annual incidence reaches up to 500 cases/100 , 000 population . Based on data of leptospirosis surveillance in New Caledonia , serogroup Icterohaemorrhagiae is the dominant serogroup involved in ca . 60% of human cases . Other serogroups involved in human leptospirosis include Pyrogenes ( 18–25% ) , Ballum , Australis and Pomona . Interestingly , the New Caledonian L . interrogans serovar Pyrogenes was formerly shown to be epidemiologically related to freshwater contaminations . Therefore , human leptospirosis cases infected with this strain provide opportunities to investigate the persistence and survival of pathogenic Leptospira in natural habitats . The purpose of the present study was to assess the presence of pathogenic leptospires in environmental samples and to estimate their viability over time . Using a TaqMan-based real time quantitative polymerase chain reaction , we screened 73 environmental samples retrieved from 4 suspected environmental infection sites for the pathogen-specific leptospiral gene lipL-32 . This study found that a large proportion of soil samples were positive for pathogenic leptospiral DNA , suggesting that repeated exposure to Leptospira may be occurring in these high-risk areas . Herein , we report findings from retrospective investigations of environmental contaminated areas to assess the presence of pathogenic Leptospira in order to better delineate and monitor high risk areas .
Four sites were identified according to the infectious strain and the good acceptance of the project by the patients and custom chiefdom ( Kaala-Gomen , Koné , Touho ( 2 sites ) , Fig 1 ) . All four study sites were within Melanesian tribes , where many outdoor activities are part of the everyday life , including fishing and bathing in freshwater streams , maintenance of backyard pig pens , hunting ( deer and wild hogs ) . In addition , two extra sites where L . interrogans Pyrogenes was known to have been involved in former cases but where no recent contamination were reported were chosen as control sites and investigated according to the same sampling procedure . Most of the investigated sites were located in the North province of the main island where climate is sub-tropical and oceanic with a hot and rainy season from December to March ( average temperature 28°C ) and a cooler season from June to September ( average temperature 20°C ) . Annual cumulative rainfall is 2400 mm on average but can range from 1460 mm to 3550 mm . Daily rainfall data for each site were obtained from the Météo France free online public database , using the nearest meteorological station for each study site . Environmental investigations were started 6 to 10 weeks after the supposed infection date . Between March and June 2016 , a total of 73 environmental samples were collected: 10 water samples , 52 soil samples and 11 other samples ( vegetal floating debris , algae ) were analyzed . Water and soil sample collections were carried out as follow: For water samples , 10 mL of subsurface water ( stream or river ) were collected at a 10–30 cm depth every 10 meters , alongside the water body directly into 15-mL sterile Falcon tubes , stored on ice and transported to the laboratory . For soil samples , approximately 50 g topsoil was collected from river banks ( from 10 cm below to 1 meter above water level ) in shaded areas using a core drilling ( 3 cm large by 5–7 cm height ) . Each soil sample was immediately placed into a 50-mL sterile Falcon tube . Water quality and environmental parameters were collected at the time of sampling ( apparent meteorological and hydrological conditions , presence of iridescences , debris , foam or stagnant fludge , water color , clarity , turbidity , salinity , temperature , dissolved oxygen , pH , UV radiation , altitude ) . The location of sampling sites was taken with a Garmin GPS . All samples were transported to the laboratory and processed within 48 hours of collection . Each water sample ( 10 mL ) was centrifuged at 8000 × g for 10 min . The pellets were resuspended to a total volume of 200 μL in the original water and immediately lysed to begin the extraction process using a commercial kit ( QIAamp DNA Mini Kit , Qiagen , Australia ) according to the manufacturer’s instructions . DNA elution was performed with 50 μL of buffer AE . The quantity of DNA was measured by NanoDrop ( Thermo Fisher Scientific ) . Soil samples were submitted to DNA extraction using the PowerSoil DNA Isolation kit ( MO BIO ) , shown in preliminary experiments to be the most efficient to extract leptospiral DNA from New Caledonian soils . Briefly , 250 mg of soil is poured in a PowerSoil bead tube before addition of 60μL suspension Buffer C1 . This suspension is shaken for 5 minutes at 2 , 000 rpm using a MagNA Lyser ( Roche ) . The supernatant is lysed at 4°C for 5 min with 250 μL lysis buffer C2 . Up to 600 μL of supernatant is transferred in a new tube before addition of 200 μL of Inhibitory Removal Technology solution C3 before incubation at 4°C for 5 min . This step is essential for the final DNA quality as it allows the cationic flocculation of humic substances which usually account for low DNA recovery and qPCR inhibition . Up to 750 μL of the supernatant is transferred in a new tube and gently mixed with 1 , 200 μL of DNA binding solution C4 prior to be loaded into a Spin Filter and centrifuged at 10 , 000x g for 1 minute at room temperature . After washing the precipitated DNA with 500 μL of wash buffer C5 through the spin filter membrane , the DNA is eluted with 100 μL elution buffer C6 . Soil and water samples were tested for the presence of pathogenic Leptospira DNA using the real-time PCR targeting lipL-32 [15] . The reactions were performed in a final volume of 20 μL containing 1X LightCycler 480 probes Master ( Roche Applied Science , New Zealand ) , 0 . 4 μM each primer and 0 . 13 μM probe , and 2 μL template DNA . The cycling conditions were as described in the original publication in a LightCycler 480 ( Roche Applied Science ) [15] . Samples with a positive lipL-32 qPCR were investigated with BLU-V Viability PMA Kit ( Qiagen ) to evaluate the presence of viable pathogenic leptospires , except for site 1 . Briefly , 5 g of soil were gently resuspended in 5 mL of 1X Phosphate Buffered Saline and let to settle down for 1 hour . Then 100 μL of this soil suspension supernatant was mixed with 2 μL of propidium monoazide ( PMA; 50 μM final concentration ) in light-transparent 1 . 5 mL microcentrifuge tubes . Following a 10 min incubation in the dark , samples were exposed for 10 min to a 3-watt LED light ( 460-470 nm ) with gentle homogenization every 2 minutes . The sample tubes were laid horizontally under the light source to ensure optimal PMA/DNA cross-linking , thus avoiding false positive results . In order to test the efficiency of PMA treatment of membrane-compromised bacterial cells , duplicate tubes of the same soil solution supernatant were heated at 80°C for 10 min . The heat-treated samples were then cooled to room temperature before PMA addition , incubation and photoactivation . In addition , a control tube without PMA was included to determine the presence of total pathogenic Leptospira ( both dead and live ) in the soil sample . After photoinduced cross-linking , samples were treated for DNA isolation using QIAamp DNA Mini kit ( Qiagen ) . The corresponding DNA extracts were used as templates for qPCR targeting the lfb-1 gene [17] in order to subsequently phylogenetically identify the viable pathogenic Leptospira present in the sample . This qPCR was run on a LightCycler LC 2 . 0 using the LightCycler FastStart DNA Master SYBR Green I kit ( Roche Applied Science , New Zealand ) as described before [17] . The lfb-1 sequence polymorphism was used as a molecular phylogenetic target to link the identity of environmental leptospiral sequences to the corresponding human infecting strain . Amplified lfb-1 DNA products obtained from environmental samples were identified by DNA sequencing . The amplicons were purified using a DNA purification kit ( Qiagen , Australia ) and sequenced directly as described before [16] . The resulting DNA sequence data were compared with sequences retrieved from the patient’s sample and with the GenBank database using the BLAST algorithm . The sequences obtained in this research were deposited in GenBank under the Accession Numbers: KY052025; KY052026; KY052027; KY052028; KY052029; KY052030; KY052031; KY052032; KY052033; KY052034; KY052035; KY052036; KY052037; KY052038; KY052039; KY052040 .
Located in the north of New Caledonia , in places were leptospirosis is endemic , 4 different sites distributed over 3 municipalities were investigated ( Fig 1 ) . For sites 1 , 2 and 4 , the infections supposedly occurred on February 2nd , during the same heavy rain event which hit New Caledonia the same day ( S1 Fig ) . All patients were swimming in a freshwater stream when the rain started to fall and all 3 reported an increase of the water flow and a change in water color and turbidity in their respective bathing sites . For site 3 , the patient probably got infected on February 9th when fishing freshwater shrimp using a mask and snorkel . For site 1 , only one investigation was performed 6 weeks after the contamination event . Sites 2 , 3 and 4 were investigated twice with 7 weeks ( site 3 and 4 ) or 10 weeks ( site 2 ) between investigations ( S1 Fig ) . The overall results for detection of pathogenic leptospires from these 4 sites are summarized in Table 1 . Interestingly , of the 10 water samples collected , none were positive for the presence of pathogenic Leptospira DNA by qPCR , either at the early or late investigation time point . Contrarily , soil samples were mostly positive: 58% of soil samples ( 30/52 ) were positive using lipL-32 qPCR . It is worth to note that among soil samples investigated , we were able to amplify pathogenic Leptospira DNA from the river bank up to 1 meter above the water level . In such a core soil sample , DNA from pathogenic Leptospira was amplified from all 1-cm thick slices down to a 5-cm depth . In addition , 2 samples mostly made of benthic algae collected on the bottom of the streams were also positive using lipL-32-qPCR . Despite a decreasing number of leptospiral DNA-positive soil samples in sites 3 and 4 , we still successfully detected pathogenic Leptospira DNA in the late samples collected 4 months after the index infection event , although the qPCR Cycle Thresholds ( Ct ) slightly increased ( S1 Fig ) . In contrast , in the two control sites where L . interrogans Pyrogenes was known to have been involved in human cases in previous years but without recent contamination reported ( > 1 year ) , none of the samples collected was positive . For each soil sample positive for lipL-32 by qPCR , we further investigated the genotype of these pathogenic leptospires using the lfb-1 phylogenetic scheme used for patients . Viability-PCR ( v-PCR ) and qPCR targeting lfb-1 using the v-PCR treated DNA as a matrix were performed subsequently when possible ( for sites 2–4 ) . The overall results for v-PCR and lfb-1-derived phylogenetic analysis from these 6 sites are summarized in Table 2 . In all 4 sites investigated , we were able to amplify DNA from L . interrogans Pyrogenes , respectively at 6; 9; 9 and 10 weeks post-infection ( WPI ) for site 1 , 2 , 3 and 4 . To further assess if this L . interrogans Pyrogenes DNA derived from live cells , we performed a viability-PCR ( for sites 2–4 ) . Two sites ( 2 and 3 ) out of the 3 investigated were positive for v-PCR and phylogenetic analyses of the amplified DNA matched to L . interrogans Pyrogenes . Interestingly , a lfb-1 sequence identical to a L . interrogans from serogroup Australis , involved in other human cases in New Caledonia , was also detected in site 2 , concomitantly with Pyrogenes . v-PCR was also positive in site 4 , but the phylogenetic analysis of the amplified lfb-1 sequences did not match the infecting strain nor any other reported isolate ( except one sequence displaying 96% identity with L . kmetyi; Fig 2 ) . To clarify whether the pathogenic leptospires could be detected over a longer period , soil samples were collected again 19 ( site 2 ) , 16 ( site 3 ) or 17 WPI ( site 4 ) . All the samples investigated from these 3 sites were negative using v-PCR . However , we were still able to amplify a few lfb-1 sequences using direct qPCR for site 3 and 4 . Phylogenetic analysis of these lfb-1 sequences appeared to differ from any known strain or species , though some were similar to those amplified during our first investigation ( Fig 2 ) . Finally , it is interesting to note that temporal analysis of our results seems to highlight dynamic changes of the pathogenic leptospires in environmental sites . Indeed , when sequences identical to L . interrogans Pyrogenes or Australis were found during our first investigation , they were either not re-detected ( site 2 and 4 ) or substituted by other unknown pathogenic leptospires upon our 2nd investigation ( site 3 ) .
Infected mammals by shedding large amounts of virulent leptospires in their urine , massively contaminate their surrounding environment [5 , 18–20] . These pathogens eventually get drained in freshwater systems upon heavy rain episodes . This dispersion through freshwater not only participates to substantial contamination of large areas but also brings the threat right in human influence area . As environmental contamination is the major source of human leptospirosis , we attempted in this study , to evidence the presence of virulent leptospires in natural habitats in New Caledonia . Using molecular-based methods , we investigated the presence and viability of pathogenic leptospires in area believed to be contamination sites . Although the use of qPCR is becoming frequent for diagnostic purpose [14] , the use of this technique on environmental samples is not commonplace , mainly because of the presence of inhibitors impairing qPCR efficiency [13] . We applied this methodology to complex environmental samples from places selected as putatively involved in human cases and we successfully amplified pathogenic Leptospira DNA in all the sites investigated . Phylogenetic analysis based on the lfb-1 gene sequence successfully linked the identity of environmental leptospiral sequences to the corresponding human cases . More importantly , we demonstrated by viability qPCR that these pathogenic leptospires were viable and probably persisted in infection sites weeks after the contamination event . Notwithstanding that little is known regarding the mechanisms regulating the persistence of pathogenic leptospires in natural habitats , outside a mammalian host , general agreement in the scientific community agrees on the fact that pathogenic Leptospira can survive for long periods in freshwater [7] and a few studies exploring the survival of Leptospira in soils successfully reported re-isolation of the same Leptospira isolate 5 months later [21] . Our results show , that when performing retrospective investigations , pathogenic leptospires could be evidenced only in soils samples , up to 4 months after the index contamination . Though repeated contaminations from an animal source might occur , our results strongly suggest a prolonged survival in river banks and soils . Moreover , to our knowledge , this study reports the first successful viability-PCR performed on Leptospira from complex soil samples in combination with molecular-based typing to identify environmental leptospires involved in human cases . We formally established that up to 9 weeks after infection , the pathogenic Leptospira strain involved in human cases was viable in environmental soil samples , and thus potentially infectious , suggesting an ongoing risk for humans . Using samples collected 4 months after the contamination event , we were not able to evidence the presence of this particular virulent strain , therefore suggesting a decrease in Leptospira viability over time , a hypothesis supported by higher Ct values in qPCR from late samples ( S1 Fig ) . These 2nd investigations were performed during the cool season , also assumed to be detrimental for Leptospira survival . Whether this change in temperature contributed to the decrease in environmental contamination remains unknown but would be in good agreement with empirical knowledge as well as the experimental results reported by Andre-Fontaine and collaborators [7] . Interestingly , we have not evidenced pathogenic leptospires in any of our water samples , contrasting with previous observations [22 , 23] . Following patients’ interviews , we have investigated the suspected flowing water bodies ( streams and rivers ) at their normal flow rate and weeks after the index contamination event . Oppositely , stagnant water sources ( gutters , wells , puddles , reservoirs ) considered in other studies [22 , 23] were not mentioned in the interviews and therefore were not investigated . In addition , we have processed a relatively small volume of water for DNA extraction ( 10 mL ) , contrasting with larger volumes ( 50–1 , 000 mL ) in other studies . Lastly , we also have investigated a smaller number of water samples compared with soils . Taken together , these facts may explain the differences observed . The detection ( or absence of detection ) of live L . interrogans Pyrogenes after long periods should be interpreted with caution because of possible limitations of the v-PCR methodology , especially in environmental samples [24] . Although selective nucleic acid intercalating dyes , like propidium monoazide used in this study , represent one of the most successful recent approaches to detect viable cells ( as defined by an intact cell membrane ) by PCR and have been effectively evaluated in different microorganisms , major drawbacks have also been reported [24] . When applied to complex environmental samples the dye may be able to penetrate into viable or reversibly damaged cells , leading to false negative results . Further , bacteria might not all be transferred from their substratum to the supernatant or be damaged during the initial steps of the v-PCR protocol . Considering that Leptospira concentrations in our samples were low , v-PCR most probably underestimated bacterial viability in our experimental procedure . Interestingly , the control sites , defined by the former presence of L . interrogans Pyrogenes but without human contamination reported over more than one year , showed no positive sample for lipL-32 qPCR detection . This not only suggests that the index patients actually got infected in the sites investigated , but also that targeting human contamination areas is a valuable strategy to properly identify Leptospira-contaminated areas , notably for research purpose . Dynamic changes in Leptospira population in environmental samples seem to have occurred over the time course of this study . While the infecting L . interrogans Pyrogenes could only be detected by qPCR during the first investigation , our study revealed the presence of other pathogenic Leptospira DNA not associated to any known species ( site 4 ) . In sites 2 and 3 , L . interrogans Pyrogenes was detected alongside with other pathogenic Leptospira spp . It is well known that microorganisms can cooperate in complex assemblages to better exploit nutritional resources and resist to stressful environmental conditions . Because leptospires are thought to be highly susceptible to adverse environmental stresses , they could have promoted a unique microbial interaction , by which leptospires would successfully survive and persist in the environment . This emerging idea has been highlighted by the recent discovery of biofilm produced by pathogenic leptospires [25 , 26] . Whether multispecific biofilms either produced by Leptospira spp . or formed by other environmental bacteria and providing shelter to leptospires , might be present in natural habitats , contribute to the persistence and allow long-term survival of pathogenic leptospires in nutrient-poor or adverse aqueous environments deserves consideration . Recent work in other settings where leptospirosis is highly endemic supports this hypothesis [9] . Interestingly , during the flooding event which occurred on February 2nd , people who got infected at sites 1 , 2 and 3 were bathing with at least 2 other persons who were exposed similarly to the Leptospira environmental risk , but did not develop a clinical form of leptospirosis . This observed low attack rate raises many questions including asymptomatic leptospirosis . A recent sero-incidence study in Brazil has shown that only a very small proportion of infections actually leads to clinical disease [27] . Overall , this study revealed that pathogenic Leptospira are widespread in river soils in places associated with recent human cases . The infecting strain was evidenced in all the investigated sites and viable leptospires were still detected 9 weeks after the contamination event . These observations are particularly interesting especially if they are analyzed in regards of daily rainfall data ( S1 Fig ) . Analysis of the daily rainfall shows that in all 4 study sites , several episodes of heavy rain occurred over the 6-month period when this study was performed . But consistently with our qPCR results these rain events have probably not been a major source of re-contamination with human threatening strains , as supported by ( i ) the fact that a similar sampling strategy failed to evidence an environmental contamination with L . interrogans Pyrogenes in late samples and ( ii ) an increase in qPCR Ct values ( S1 Fig ) , suggesting a decrease of the environmental Leptospira load over time . Therefore , it is likely that the Leptospira DNA that was detected over this 4-month study corresponded to leptospires deposited by the flooding event of February 2nd . Still , temporal investigations evidenced changes in leptospiral diversity and revealed the presence of yet unreported strains in soil samples and never evidenced in any mammal in New Caledonia despite long research , suggesting that soil might act as an environmental reservoir of pathogenic leptospires offering them a protective atmosphere . v-PCR coupled to molecular-based typing on soil samples proved effective at confirming infection sites and investigating the leptospiral risk over time . Because soil DNA extraction only uses small amounts of soil , the use of this approach for risk evaluation should consider the possibility of false negative results . Still , the assessment and quantification of the leptospiral burden in environmental samples might prove valuable to guide public health interventions . To help expand the current knowledge about the leptospirosis environmental cycle and the spatial and temporal distribution of leptospires in the environment , further studies will also characterize the physicochemical characteristics of soils shown to support or oppositely compromise the survival of pathogenic leptospires . Furthermore , determination of the environmental burden may help inform health authorities before adopting preventive measures such as access restrictions to contaminated areas during heavy rainfall events . Finally , evaluation of the environmental leptospiral load through quantitative methods can be a useful method to monitor high risk areas and help protect local populations , but also to discover an unexplored biodiversity of pathogenic leptospires . | Leptospirosis is an emerging zoonotic disease caused by infection with pathogenic strains of Leptospira . Most human infections arise from environmental exposure to contaminated freshwater environments or watered soils where pathogenic Leptospira are considered as able to survive for prolonged periods . Therefore , a good understanding of Leptospira survival strategy in the environment is a key step to identifying crucial factors amenable to interventions and public health actions to lower leptospirosis burden . In this study , we investigated the environmental presence and survival of pathogenic leptospires in areas where recent human leptospirosis cases had been reported . Although detection of Leptospira from complex environmental samples is difficult , we successfully detected the presence of pathogenic Leptospira in soils of suspected infection sites . In addition , we showed that these pathogenic leptospires were alive and present in soils several weeks after the infecting event . Typing of leptospiral DNA retrieved from the environment revealed identities between environmental pathogenic Leptospira and the causative strains involved in human leptospirosis index cases . Interestingly , we also identified yet unreported genotypes . Altogether , our work illustrates the potential of quantitative molecular assays for the rapid detection and typing of viable leptospires in environmental samples , which could prove useful to assess the risk of environmental exposure . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"surface",
"water",
"leptospira",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"tropical",
"diseases",
"microbiology",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"bacteria",
"bacterial",
"pathogens"... | 2017 | Seeking the environmental source of Leptospirosis reveals durable bacterial viability in river soils |
Molecular mechanisms for the establishment of transcriptional memory are poorly understood . 5 , 6-dichloro-1-D-ribofuranosyl-benzimidazole ( DRB ) is a P-TEFb kinase inhibitor that artificially induces the poised RNA polymerase II ( RNAPII ) , thereby manifesting intermediate steps for the establishment of transcriptional activation . Here , using genetics and DRB , we show that mammalian Absent , small , or homeotic discs 1-like ( Ash1l ) , a member of the trithorax group proteins , methylates Lys36 of histone H3 to promote the establishment of Hox gene expression by counteracting Polycomb silencing . Importantly , we found that Ash1l-dependent Lys36 di- , tri-methylation of histone H3 in a coding region and exclusion of Polycomb group proteins occur independently of transcriptional elongation in embryonic stem ( ES ) cells , although both were previously thought to be consequences of transcription . Genome-wide analyses of histone H3 Lys36 methylation under DRB treatment have suggested that binding of the retinoic acid receptor ( RAR ) to a certain genomic region promotes trimethylation in the RAR-associated gene independent of its ongoing transcription . Moreover , DRB treatment unveils a parallel response between Lys36 methylation of histone H3 and occupancy of either Tip60 or Mof in a region-dependent manner . We also found that Brg1 is another key player involved in the response . Our results uncover a novel regulatory cascade orchestrated by Ash1l with RAR and provide insights into mechanisms underlying the establishment of the transcriptional activation that counteracts Polycomb silencing .
Studies on the regulation of transcriptional memory are challenging . Conceptually , the regulation consists of two phases: establishment and maintenance . Molecular mechanisms for the maintenance of the memory are relatively well understood compared with those for the establishment of memory . Indeed , how the establishment of transcriptional activation occurs is largely unknown because it has been difficult to distinguish mechanisms for establishment from those for maintenance , presumably due to functional redundancies and spatial and temporal overlap between them . Moreover , if transient regulation is involved during the establishment phase , it is extremely difficult to tease apart and analyze the respective underlying mechanisms . For the establishment of transcriptional activation of developmentally regulated genes in stem cells , we know that the poised RNAPII should be released from pausing in the promoter-proximal coding region , as occurs in response to various microenvironmental cues [1] , and that the associated chromatin should be kept competent for transcription by RNAPII throughout a coding region . P-TEFb , a cyclin-dependent kinase complex , plays a pivotal role in the RNAPII pause release by alleviating the repressive effects of DRB sensitivity-inducing factor ( DSIF ) and negative elongation factor ( NELF ) , and by phosphorylating the Ser2 residue of the carboxyl-terminal domain of RNAPII [2]–[5] . In addition to recruitment of P-TEFb , it has been proposed that recruitment of a certain chromatin remodeling factor is also crucial to the release of paused RNAPII [3] , which appears to be situated adjacent to the first nucleosome downstream of the transcription start site [6] . Thus , it seems that at least two independent mechanisms are required to trigger productive transcriptional elongation . It is conceivable that these mechanisms are engaged throughout the coding region to maintain active gene expression . In addition , the activation of Polycomb group-target genes further requires several counteracting mechanisms against the Polycomb repressive complexes ( PRCs ) [7] . However , these mechanisms underlying the establishment of transcriptional memory and how these mechanisms are orchestrated remain elusive . Ash1l is the mammalian equivalent of the Drosophila Ash1 protein . Although Ash1 is one of the first identified members of the trithorax group proteins [7] , both the Ash1- and Ash1l-containing complexes remain uncharacterized . Both Ash1 and Ash1l are localized in chromatin and have been identified specifically in promoter-proximal coding regions of a number of active genes [8] , [9] , suggesting a role during an early step of transcriptional elongation . Additionally , artificial tethering of Ash1 to chromatin containing a reporter gene results in gene activation in a SET domain-dependent manner [10] . These results suggest that Ash1 is an epigenetic activator found in an “ON” state of target genes , although the underlying mechanism of its action remains unknown . Like a number of SET domain-containing proteins , both Ash1 and Ash1l possess histone lysine methyltransferase activity . However , it is currently controversial as to which lysine residue is targeted in vivo , although recent reports have suggested that Lys4 and Lys36 of histone H3 are the most plausible candidates [8]–[11] . So far , Lys4 is widely believed as a target residue due to several lines of in vivo evidence [8]–[10] , compared with only one for Lys36 [12] . It is consistent with the activator function of Ash1 [10] , while Lys36 methylation ( Lys36me ) has been shown to occur as a consequence of transcription [13]–[15] . However , it should be noted that an analysis of the enzymatic activities of Ash1 and Ash1l in vivo has been difficult due to their physical and functional interactions with other enzymes that methylate Lys4 [16] , [17] and also due to the redundancy of enzymes methylating Lys36 , such as Setd2 , Nsd1 , Nsd2 and Nsd3 in mammals . Moreover , expression of the full-length recombinant Ash1l and detection of the endogenous Ash1l by immunoblot has been as yet unsuccessful , thereby making analyses more challenging . To elucidate the molecular mechanism of the Ash1l-mediated establishment of gene expression , we developed a knock-in allele expressing mutant of Ash1l without part of the SET domain . Using the mutant ES cells , we show that Ash1l methylates Lys36 of histone H3 both in vitro and in vivo . Importantly , using DRB , a P-TEFb kinase inhibitor that blocks productive transcriptional elongation , we show that the methylation by Ash1l and its effect on PRC exclusion occur independently of productive transcriptional elongation . In particular , an accumulation of Lys36me3 in RAR-associated genes independent of transcriptional elongation implicates a certain special function for establishing transcriptional activation of Polycomb group-target genes . Moreover , the broad chromatin domains carrying Lys36me were co-regulated with the Tip60 or Mof complexes in a region-dependent manner , which in turn acetylate the Lys16 of histone H4 . We further investigated a mechanism for the promotion of gene expression by Ash1l , and found an Ash1l-dependent association with Brg1 , another founding member of the trithorax group proteins . These molecular data in ES cells are supported by expression patterns of Hox genes and skeletal phenotypes in Ash1l mutant mice . Thus , through genetic and biochemical analyses of Ash1l , we have elucidated a novel cascade of interplays from Ash1l to Brg1 , which ultimately promotes chromatin reprogramming that counteracts Polycomb silencing .
To explore a function of the methyltransferase activity of mammalian Ash1l , we generated knock-in mice and ES cells expressing mutant Ash1l containing a short in-frame deletion within the AWS-SET domain ( represented by ΔSET , Figure 1A ) . Mouse Ash1l mRNA was ubiquitously expressed in embryos , while it was relatively enriched in the adult brain ( Figures S1A , S1B and data not shown ) . Expression of Ash1l mRNA was not affected in ΔSET embryos and ES cells ( Figures 1B and 1C ) . Homozygotes were viable and fertile ( Table S1 ) . Since Ash1l is a member of the trithorax group proteins that regulate transcriptional activation of Hox genes , and retinoic acid ( RA ) is known to induce expression of Hox genes in ES cells , a role for Ash1l methyltransferase activity in RA-induced Hox gene expression was investigated in differentiating ES cells . A culture protocol is shown in Figure 1D . Here , we analyzed representative RA-responsive Hox genes , Hoxb4 , and Hoxd4 . Using a series of RA concentrations , we found that expression levels of Hoxb4 and Hoxd4 mRNAs were reduced in ΔSET ES cells ( Figure 1E ) . Specifically , the threshold concentration of RA required to trigger the Hoxd4 mRNA expression was significantly increased in ΔSET ES cells compared with that in wild-type cells ( Figure 1F ) . Moreover , the activation of the Hoxd4 mRNA expression was relatively slow in ΔSET ES cells ( Figure 1G ) . These results indicate that the methyltransferase activity of Ash1l is necessary for an appropriate response to RA . We further performed gene expression analyses of RA-treated differentiating ES cells by RNA-sequencing ( RNA-Seq ) to determine if RA-responsive genes were generally affected in ΔSET ES cells . Among 14 , 255 annotated genes , there were 543 genes that were highly up-regulated by RA treatment ( more than 5-fold ) . Among those 543 genes , we found 152 genes ( 28 . 0% ) in ΔSET ES cells showing impaired responses to RA ( more than 2-fold decrease compared with wild-type cells , Figure 2A and Table S2 ) , in which several Wnt and Hox family genes appeared to be affected ( Figure S2 ) . As expected , gene ontology analysis of the 152 genes revealed that biological functions of Ash1l were highly related to body pattern formation during development ( Figure 2B ) . The status of chromatin signatures in ES cells can be classified in terms of the presence of Lys4me3 or Lys27me3 in histone H3 polypeptides: Lys4me alone , Bivalent ( positive for both Lys4me3 and Lys27me3 ) , Lys27me alone and None ( negative for both Lys4me3 and Lys27me3 ) [18] . Interestingly , the 152 RA-responsive and ΔSET-impaired genes were significantly enriched in a group positive for Lys27me3 ( Figure 2C ) and further enriched especially in the Bivalent gene group ( Figure 2D ) , suggesting that the methyltransferase activity of Ash1l counteracts Polycomb silencing upon activation of developmentally regulated genes . Chromatin immunoprecipitation-sequencing ( ChIP-Seq ) analysis revealed that Ash1l was present in more than 30% of those affected genes ( Figure 2E ) . RNA-Seq analysis of undifferentiated ES cells showed that expression levels of a majority of marker genes , including core stem cell markers , were unchanged , while some endoderm markers were moderately up-regulated in ΔSET ES cells ( Figure S3A ) . RA treatment of ES cells further enhanced the up-regulation ( Figure S3B ) . Comprehensive analysis of RNA expression levels of 14 , 255 annotated genes revealed that 57 genes were down-regulated more than 2-fold in undifferentiated ΔSET ES cells , while 59 genes were up-regulated ( Table S3 ) . Several microRNAs and Snord family genes were highly dys-regulated in ΔSET ES cells , although the impact of the methyltransferase activity of Ash1l on these genes in vivo remains unclear . While this manuscript was under preparation , a report describing the methyltransferase activity of Ash1l for Hox gene repression was published online [19] . Reason for discrepancy between our results and theirs is currently unknown . However , conditions of basic materials and methods may affect each result: knock-in mutant ES cells and mice for ours , exogenous transient over-expression in K562 cells for theirs . To elucidate an underlying mechanism involved in the regulation of gene expression by Ash1l , we examined the biochemical activity of Ash1l in vitro . We employed a bacterially expressed GST-fusion Ash1l protein ( Figure 3A ) since it has no associated protein that methylates histone polypeptides . As shown in Figure 3B , wild-type Ash1l ( NF-WT ) methylated histone H3 in nucleosomes , but did not methylate free histone octamers or ( H3-H4 ) 2 tetramers bound to DNA . Addition of H2A-H2B dimers to the ( H3-H4 ) 2 tetramers on DNA in the reaction mixture failed to replicate the methylation activity [K . N . unpublished observation] , and an amino acid substitution within the core catalytic part of the SET domain ( NF-H2213K ) abolished it . Additionally , wild-type Ash1l methylated wild-type histone H3 and its Lys4-to-Ala ( K4A ) mutant to a similar extent , but not its Lys36-to-Ala ( K36A ) mutant ( Figure 3C ) . These results clearly indicate that Ash1l specifically methylates Lys36 of histone H3 and presumably recognizes the preinstalled H2A-H2B dimer in a nucleosome to target Lys36 . Furthermore , recombinant Ash1l carrying a deletion at the N-terminal flanking region of the AWS domain ( ΔNF-WT ) was catalytically inactive , indicating that this region is necessary for optimum methyltransferase activity . We next examined whether our in vitro results could be recapitulated in an in vivo setting . Histone modification patterns and Ash1l occupancy in Hoxd4 in differentiating ES cells were analyzed by ChIP assays . PCR primer-pairs were set as shown in Figure 3D ( Table S4 for sequences ) . A promoter-proximal coding region of Gapdh was also investigated as a constitutively active control . Hoxd4 chromatin has been reported to be bivalent in undifferentiated ES cells [18] . Consistent with the report , even in the absence of RA , peaks of histone H3 Lys4me2 and me3 ( Lys4me2/3 ) were clearly detected in a promoter-proximal coding region of Hoxd4 , while the levels of histone H3 Lys36me2 and me3 ( Lys36me2/3 ) were rudimentary ( Figure 3E ) . Surprisingly , Ash1l was clearly present in Hoxd4 even in the absence of RA ( Figure 3E ) . We also observed considerable amounts of Ash1l in the promoter-proximal coding region of Gapdh . However , we observed no difference between wild-type and ΔSET ES cells in these ChIP assays . Following RA treatment , while the levels of Lys36me2/3 were increased in Hoxd4 coding regions in wild-type cells , this was not observed in ΔSET ES cells ( Figure 3F ) . There was no significant change in the levels of Lys4me2/3 of Hoxd4 in ΔSET ES cells . We found that the levels of Ash1l in coding regions of Hoxd4 and Gapdh were maintained after RA treatment . We observed no difference in the levels of Ash1l between wild-type and ΔSET ES cells . These results indicate that Ash1l specifically methylates Lys36 in vivo , as expected from our in vitro results , and suggest that the enzymatic activity of Ash1l is activated upon the addition of RA . Since Ash1l is a dimethylase [11] , the observation that there was no increased Lys36me3 in ΔSET ES cells may be a consequence of an impaired Lys36me2-platform that is required as a substrate for a certain trimethylase . Based on several studies conducted in yeast , the presence of Lys36me in a coding region is widely believed to be a consequence of transcriptional elongation [13]–[15] and to function to recruit the histone deacetylase complex [20] , [21] . However , our results suggested a novel hypothesis: Lys36me by Ash1l in a coding region occurs during the establishment of Hox gene activation to promote a proper response to RA ( see Figures 1F and 1G ) . Therefore , in this case , the Lys36me should be independent of the productive transcriptional elongation . To test our hypothesis , we employed DRB , which reversibly blocks productive transcriptional elongation by inhibiting the kinase activity of P-TEFb . DRB artificially creates the poised RNAPII , closely mimicking the promoter-proximally paused RNAPII in a gene demonstrating bivalent chromatin in ES cells , thereby manifesting intermediate steps for the establishment of transcriptional activation . If Lys36me by Ash1l is independent of the RNAPII Ser2p or transcriptional elongation , then DRB would not affect methylation levels . In the next experiments , DRB was added to differentiating ES cells during RA treatment and ChIP assays were performed . Two Ash1l-associated genes were compared: Hoxd4 , representing a Polycomb group-target gene , and Gapdh , a constitutively active gene . As shown in Figure 4A , far from a decrease in the Lys36me level , wild-type ES cells displayed a clear increase in Lys36me2/3 levels in the promoter-proximal coding region of Hoxd4 in response to DRB . Interestingly , differences in the Lys36me2/3 levels between wild-type and ΔSET ES cells were more evident in the presence of DRB . On the other hand , DRB also increased the Lys36me2/3 levels in ΔSET ES cells , albeit to a lesser extent [compare ( − ) and ( + ) in ΔSET] , suggesting that some Lys36-methylases other than Ash1l could be involved , although the identity of the enzyme recruited to the region has yet to be determined . In Gapdh , DRB increased Lys36me2 levels in wild-type cells , but DRB treatment resulted in a clear decrease in Lys36me3 levels ( Figure 4B ) , suggesting that DRB specifically affected association of a certain trimethylase that was recruited in a RNAPII Ser2p-dependent manner . Similar results were obtained for Hoxb4 and Hprt1 ( Figure S4 ) . These results indicate that Lys36me2 by Ash1l and other dimethylases occurs independently of RNAPII Ser2p . However , whether Lys36me3 occurs independently of RNAPII Ser2p is gene-dependent . Having established that Hoxd4 and Gapdh were DRB-responsive genes in that Lys36me2 ( /3 ) was increased in response to DRB , we analyzed if the observed response was applicable genome-wide . First , immunoblot analyses for bulk Lys36me2 and Lys36me3 levels were performed ( Figure 4C ) . In the presence of RA , we found the bulk Lys36me2 levels in wild-type and ΔSET ES cells were similarly increased in response to DRB , while those of Lys36me3 were not , and instead decreased . Next , we performed ChIP-Seq analyses to obtain genome-wide profiles of Lys36me2/3 in response to DRB . Distributions of Lys36me2/3 relative to a metagene show a clear difference between Lys36me2 and me3 in response to DRB: Lys36me3 levels were decreased in isolated , entire regions , while in contrast , Lys36me2 levels were increased in most of the regions ( Figure 4D ) . Thus , having observed that Lys36me3 of Hoxd4 was increased in response to DRB , here we found that its genome-wide level showed the opposite response , indicating that genes demonstrating increased Lys36me3 levels in response to DRB were minorities . Figures 4E and 4F show profiles of the representative genomic regions . In these genomic profiles , we found that the level of Lys36me2 in wild-type cells was generally unchanged in response to DRB in a broad range of regions , or rather increased in some parts , while that of Lys36me3 was decreased in most regions . However , we also found that substantial levels of Lys36me3 remained in scattered regions , including some inter-genic regions . Interestingly , quite a few such regions were found in the vicinity of RAR binding sites [22] , implicating a functional relationship between Lys36me3 in response to DRB and RA signaling . Indeed , the numbers of RAR-associated genes were significantly underrepresented in a gene group with decreased levels of Lys36me3 in response to DRB ( Figure 4G , ΔLog2<−0 . 6 ) . Most strikingly , RAR binding sites showed a major peak in the Lys36me3 ChIP-Seq read density plot ( Figure 4H ) . Additionally , in B16 cells , which express Hoxd4 constitutively without the addition of RA , we observed only a small increase in Lys36me3 levels in response to DRB ( Figure S5 ) . These results further support our proposed relationship . Taken together , our results indicate that Lys36me2 by Ash1l and other dimethylases occurs independently of RNAPII Ser2p in a large number of genomic regions . However , whether Lys36me3 occurs independently of RNAPII Ser2p is context-dependent , and at least in a portion of genomic regions , RAR may play a substantial role for maintaining Lys36me2/3 levels . In the above experiments , it is possible that past productive transcriptional elongation had left unknown traces on the transcribed chromatin , which in turn was targeted by several Lys36-methylases including Ash1l , although recruitment of the methylases is independent of RNAPII Ser2p . Moreover , it remains unclear whether the activation of the enzymatic activity of Ash1l requires the productive transcriptional elongation . Therefore , in the next experiments , to block the productive transcriptional elongation completely , we used DRB prior to administering RA . Specifically , DRB was added 1 hour before the addition of RA , and then the cells were cultured for another 16 hours in the presence of RA ( Figure 5A ) . Under these conditions , Hoxd4 mRNA was not increased at all from the basal level that was observed in undifferentiated ES cells , indicating that DRB blocked the productive transcriptional elongation completely ( Figure 5B ) . As shown in Figure 5C , while wild-type ES cells displayed mild increases in Lys36me2/3 levels in the promoter-proximal coding region of Hoxd4 in response to DRB , ΔSET ES cells did not , resulting in clear differences between wild-type and ΔSET ES cells in the presence of DRB . The results indicate that Ash1l-dependent Lys36me2/3 in Hoxd4 occurs independently of productive transcriptional elongation during the establishment of transcriptional activation . This may be reasonable since Ash1l is preloaded on the Hoxd4 chromatin before the addition of RA ( see Figure 3E ) . Can transcription-independent Lys36me by Ash1l counteract Polycomb silencing ? A previous report showed that transcription is necessary to exclude the PRCs from local chromatin [23] . However , proof remains elusive of whether progression of RNAPII itself is the major determinant factor for the exclusion . Moreover , how the PRCs are removed upon gene induction is poorly understood . Therefore , under the same conditions , i . e . the addition of DRB prior to RA , we characterized the status of Polycomb silencing in ES cells by analyzing Suz12 ( a component of the PRC2 ) , Lys27me3 ( an enzymatic product of the PRC2 ) , ubiquitination of histone H2A ( H2Aub , an enzymatic product of the PRC1 ) , Mel18 , and Rnf2 ( components of the PRC1 ) in Hoxd4 chromatin . We found significantly higher levels of Mel18 and Rnf2 in Hoxd4 chromatin of ΔSET ES cells compared to wild-type cells in the absence and presence of DRB ( Figure 5D ) . Interestingly , wild-type and ΔSET ES cells displayed clear decreases in Mel18 and Rnf2 levels upon blocking of transcription by DRB , demonstrating anti-parallel ChIP patterns against those of Lys36me2/3 ( compare Figures 4A , 5C and 5D ) . Similar results were obtained for a distal coding region of Hoxd4 ( Figure S6C ) . Suz12 and H2Aub levels showed more rapid and clear decreases in response to RA . However , differences between wild-type and ΔSET ES cells in the occupancies of Suz12 and H2Aub were unclear , suggesting that there was an Ash1l-independent pathway to exclude these molecules . Lys27me3 levels displayed only a marginal response to both DRB and RA under these conditions ( Figure 5D ) . However , we observed a clear decrease in Lys27me3 levels after a longer induction by RA , in which there was a substantial difference between wild-type and ΔSET ES cells ( Figure S6D ) . In summary , these experiments showed that Suz12 , H2Aub , Mel18 and Rnf2 demonstrated relatively rapid responses to RA compared with Lys27me3 , and contradicting a previous notion , their exclusion was not dependent on transcriptional elongation . Importantly , we found that exclusion of Mel18 and Rnf2 from chromatin upon RA induction was specifically impaired by loss of the methyltransferase activity of Ash1l , suggesting a negative relationship between PRC1 chromatin association and Ash1l activity . Although ΔSET ES cells displayed mild decreases in RNAPII Ser2-phosphorylation ( Ser2p ) levels in the coding regions of Hoxd4 , the decreased levels of Lys36me did not affect the basic status of RNAPII for the most part ( Figures S7A–S7C ) . Similar results were obtained even for a relatively larger gene , Wnt6 ( Figures S7D and S7E ) . These results suggest that the methyltransferase activity of Ash1l mainly contributes to promoting the anti-Polycomb silencing function rather than the activation of RNAPII directly . Having established that Lys36me in Hoxd4 occurs independently of productive transcriptional elongation and that DRB enhances the difference between wild-type and ΔSET ES cells , we next asked how the Lys36me facilitates transcriptional elongation . Given that certain histone acetylations have more direct effects on activation of transcription , ChIP assays were performed to analyze the effects of Lys36me on representative histone acetylations , such as Lys9/14 acetylation of H3 ( Lys9/14ac ) and Lys16 acetylation of H4 ( Lys16ac ) . In all subsequent experiments , when necessary , DRB was added during RA treatment as in Figure 4 . Interestingly , the ChIP pattern of Lys16ac in Hoxd4 was similar to that of Lys36me in that the ChIP signals were not decreased in the presence of DRB . In fact , they were increased in wild-type ES cells , and became clearly lower in ΔSET ES cells compared with wild-type cells ( Figure 6A ) , thereby revealing the effect of Lys36me by DRB . The ChIP pattern of Lys9/14ac did not resemble even slightly that of Lys36me . These results collectively indicate that Lys16ac specifically correlates with Ash1l-dependent Lys36me , both of which are independent of RNAPII Ser2p . This was consistent with a recent report conducted in Drosophila , where connections were made between Lys36me2 with dMes-4 and Lys16ac by an unknown enzyme in proximal coding regions [24] . Interestingly , in our study , similar results were also obtained for Gapdh ( Figure 6B ) , even in a further downstream distal coding region ( Figures 6C and 6D ) , suggesting that cooperative action between Lys36-methylases including Ash1l and a certain Lys16 acetyltransferase influences these histone modifications in an entire coding region independently of RNAPII Ser2p . Similar results were obtained for Hoxb4 and Hprt1 ( Figure S8 ) , suggesting that the observed parallel link is a general phenomenon . In ΔSET ES cells , the global levels of Lys36me2 and Lys16ac , but not of Lys36me3 , were found to be reduced ( Figure 6E ) , which further corroborated the ChIP results . Since we observed no significant difference in the levels of Ash1l , RNAPII , and Ser5p between wild-type and ΔSET ES cells ( see Figure 4A ) , we speculated that Lys36me contributed to the association of a certain Lys16-acetyltransferase with a coding region chromatin . We next analyzed the Mof and Tip60 complexes as these complexes preferentially acetylate Lys16 of histone H4 and are highly relevant to transcriptional activation . Furthermore , since these complexes contain chromodomain proteins ( Msl3l [25] in the Mof complex and Mrg15 [26] in the Tip60 complex ) that bind Lys36-methylated histone H3 [27] , [28] , both complexes can associate with the Lys36-methylated chromatin . The ChIP patterns of Tip60 in the promoter-proximal coding region of Hoxd4 and Gapdh were similar to those of Lys16ac , while Mof showed distinct patterns ( Figures 6F and 6G ) . However , in the distal coding region of Gapdh , the ChIP pattern of Mof was similar to that of Lys16ac ( Figure 6H ) , while the similarity in that of Tip60 became less prominent . These results suggest that both Tip60 and Mof are the enzymes that acetylate Lys16 downstream of Ash1l-dependent Lys36me and that they differentially associate with a target gene in a region-dependent manner , i . e . Tip60 in a promoter-proximal coding region and Mof in a distal coding region . The involvement of the acetyltransferase activity of Tip60 in Hoxd4 activation was further suggested by utilizing Tip60 knock-in mutant ES cells ( heterozygote ) ( Figure S9 ) . Under these conditions , Lys36me2 was likely to be affected , suggesting crosstalk between Ash1l and Tip60 . Having demonstrated functional interaction between Lys36me by Ash1l and Lys16ac by Tip60 or Mof , we then analyzed other events that the interaction influences . Among the chromatin remodeling complexes associated with gene activation , several in vitro studies suggest that the Brg1 complex is the most plausible candidate that targets Lys16ac [29] , [30] , although whether this applies in vivo remains unclear . The ChIP pattern of Brg1 in the promoter-proximal coding region of Hoxd4 was mostly similar to those of Lys36me and Lys16ac ( Figure 6I , left panel ) . We observed a similar result for Gapdh ( Figure 6I , middle panel ) , even in the distal coding region ( Figure 6I , right panel ) . Next , as the active P-TEFb complex containing both Cdk9 and Brd4 has been shown to target Lys16ac [31] , we also analyzed the occupancy of Cdk9 . However , the ChIP pattern of Cdk9 showed only a limited similarity to those of Lys36me and Lys16ac ( data not shown ) . Therefore , these results suggest that Lys36me by Ash1l contributes to Brg1 association in an entire coding region . We next examined whether our results in ES cells could be recapitulated in development of mice . Whole-mount in situ hybridization was employed to determine expression patterns of representative Hox genes in various parts of developing embryos that carry the ΔSET mutation . While the expression patterns of Hoxb4 , d4 , and a4 mRNAs were largely similar between wild-type and ΔSET embryos , the anterior boundaries of their expression domains were shifted posteriorly along the antero-posterior axis at the paraxial mesoderm in ΔSET embryos ( Figures 7 , S10A and S10B ) . Thus , consistent with the results in ES cells , these findings suggest that the methyltransferase activity of Ash1l promotes a response to a certain activating cue that triggers Hox gene expression during development . To examine whether the observed posterior shifts of the expression domains of Hoxb4 and Hoxd4 mRNAs are reflected by skeletal phenotype , we compared vertebrae of wild-type and mutant mice . Consistent with Ash1l being one of the trithorax group proteins , obvious alterations in the identities of vertebrae were observed ( Table 1 , Figures 8A and S10C–S10I ) . In particular , 42–56% of the mutant mice had cervical vertebrae affected , showing the homeotic transformation of the C2 vertebra into the C1 vertebra . These phenotypes were similar to those caused by mutations in group 4 Hox genes , since the C2-to-C1 transformation was caused by a functional loss of either Hoxb4 or Hoxd4 [32] , which support our results in ES cells and embryos . Importantly , we found that the ΔSET allele partially suppressed the C2-to-C3 transformation caused by homozygous mutations in Mel18 , indicating a role for Ash1l in anti-Polycomb silencing in vivo ( Figure 8B ) .
In contrast to prevalent notions , at least with regards to Hox gene activation , the present study has shown that both Lys36me2/3 in a coding region and the accompanying exclusion of the PRCs from the region occur independently of productive transcriptional elongation . RNA-Seq analysis revealed a significant functional relationship between Ash1l and Polycomb-regulated genes in that Ash1l-mediated Lys36me counteracts Polycomb silencing . Intriguingly , ChIP-Seq analysis has suggested that the preceding Lys36me2/3 during the establishment of Hox gene expression is applicable to a subset of RAR-associated genes . We have also uncovered a functional link among Ash1l , Tip60 , Mof , and Brg1 , which cooperatively promote Hox gene expression in response to RA . Collectively , our results reveal insights into mechanisms underlying the establishment of transcriptional memory that counteracts Polycomb silencing , which have until now been difficult to analyze by conventional methods . Here , we propose that Ash1l and RAR coordinate to orchestrate a novel regulatory cascade of chromatin reprogramming ( Figure 9 ) . The Lys36me2/3 preceding productive transcriptional elongation may directly counteract association of the PRCs in target chromatin [33] , [34] , resulting in de-repression from Polycomb silencing , likely through loosening of the compacted chromatin structure [35] . Therefore , Lys36me2/3 by Ash1l and other Lys36-methylases constitute a rate-limiting step , which may promote Lys16ac by Tip60 or Mof in a region-dependent manner . Lys16ac may lead to further loosening of the chromatin structure [29] , allowing the Brg1 chromatin remodeling complex to be associated and to promote chromatin reprogramming , presumably by further excluding the PRCs to alleviate Polycomb silencing and by remodeling nucleosomes to facilitate productive transcriptional elongation . The proposed mechanism might be also applied to transcriptional regulation in the Drosophila species , showing a correlation between Lys36me2 and Lys16ac [24] . Indeed , the consecutive regulatory steps described above might explain a previous report detailing progression of the ecdysone-induced puff 74EF in polytene chromosomes of Drosophila larvae under pretreatment with DRB [36] . However , the mechanism would not apply in yeast , in which an anti-correlation between Lys36me2 and histone H4 acetylation has been reported [20] , [21] . These observations suggest that such regulatory mechanisms are unique to metazoans . What is the significance of Lys36me3 during the establishment of transcriptional activation ? At least in a promoter-proximal coding region of Hoxd4 , we found Lys36me3 could occur independently of productive transcriptional elongation ( Figure 5C ) . An accumulation of Lys36me3 on the Lys36me2-platform may ensure de-repression from Polycomb silencing because Lys36-demethylases Kdm2a/b would not recognize Lys36me3 as a substrate [33] . The degree of Lys36-trimethylase recruitment was presumably RA-dependent as we observed only a small increment in the level of Lys36me3 in the presence of DRB in B16 cells without addition of RA ( Figure S5 ) . A predisposition to underrepresent RAR-associated genes in the “decreased” gene groups in response to DRB as well as accumulations of Lys36me2/3 around RAR binding sites further support our surmise ( Figures 4G and 4H ) . Of note , DRB clearly increased the Lys36me2/3 levels in promoter-proximal coding regions of Hoxb4/d4 in ΔSET ES cells ( Figures 4A and S4B ) . Therefore , we speculate that several Lys36-methylases , including Ash1l , play a role during the establishment of transcriptional activation in an RA-dependent manner . Consistent with this speculation , Ash1l , Nsd1 , and a major mammalian Lys36-trimethylase Setd2 , all have a nuclear receptor binding motif , LXXLL . Indeed , approximately 60% of RAR binding sites were co-occupied with Ash1l ( Figure S1G ) . Thus , it is tempting to speculate that nuclear receptor-dependent developmental programs may have similar underpinnings to the Hox genes regulator mechanisms revealed in this study . Our results suggest that a part of the function of Lys36me2/3 in Hoxd4 mRNA expression is masked after productive transcriptional elongation . Specifically , the effect of Lys36me2/3 on the association with Tip60 and Brg1 was more evident in the presence of DRB ( Figures 6F–6I ) , suggesting that this association is partially dependent on P-TEFb activity . Once the active P-TEFb complex associates with target chromatin and triggers the productive transcriptional elongation , it may have a dominant effect on the association over that of Lys36me2/3 . However , upon gene activation but before tethering of the P-TEFb complex , Lys36me2/3 may have a dominant comprehensive function , involving exclusion of the PRCs and promoting association of Tip60 and Brg1 , thereby facilitating the RA response ( Figure 9 ) . This idea is consistent with our results demonstrating that sensitivity against a certain activating cue appeared to be affected in ΔSET mice and ES cells ( Figures 1F , 7 , S10A and S10B ) . One of the important issues when studying transcription mechanisms on a chromatin template is how a dramatic change in chromatin structure occurs upon gene activation: in particular , whether the open chromatin structure is established before or after the first RNAPII travels along the template DNA [37] . So far , it is widely believed that a specially equipped RNAPII , or so-called “pioneer polymerase” , is required for the initial opening of the condensed chromatin . This special RNAPII breaks down the condensed chromatin structure into the open structure during the first transcriptional elongation , thereby ultimately creating the transcription-competent chromatin . However , the results of the present study led us to the notion that the driving forces initiated by the methyltransferase activity of Ash1l promote the establishment of the open and transcription-competent chromatin structure prior to the first productive transcriptional elongation by fully-activated RNAPII . Our hypothesis may be applied to active but non-productive bivalent genes; however , it remains unclear whether it can be applied to inactive , inducible monovalent genes . Results from whole-mount in situ hybridization analyses in Ash1l ΔSET mice ( Figures 7 , S10A and S10B ) were clearly distinct from those in mutant mice carrying a deletion in the SET domain of Mll1 , a representative Lys4-methylase belonging to the trithorax group , which displayed a normal expression boundary and an impaired maintenance of Hoxd4 mRNA expression [38] . On the other hand , the results in Ash1l ΔSET mice were similar to those in the Polycomb group mutant mice in that the mutants demonstrated shifts of expression boundaries at the paraxial mesoderm ( Mel18 in [39]; Phc1/2 in [40] ) , although directions of the shifts in Polycomb group mutant mice were opposite to those in Ash1l ΔSET mice . Collectively , these results suggest that Ash1l has a distinct function from Mll1 and directly counteracts the function of the Polycomb group proteins . Consistent with this idea , Ash1l ΔSET mice only demonstrated additive and non-synergistic phenotypes with the double-heterozygous Mll1 mutation [YY & KN , unpublished observation] , and a partial suppression in the phenotype with the Mel18 mutation ( Figure 8B ) . We also observed that Ash1l was localized in a promoter-proximal coding region ( Figures 3E , 3F and S1E ) , corroborating previous reports [8] , [9] . Bromo- , PHD- and BAH-domains in the carboxyl-terminal region of Ash1l supposedly function to restrict localization . The distribution of Ash1l in Hoxd4 was similar to those of Lys4me2/3 , and a large portion of Ash1l was co-localized with Lys4-methylated chromatin ( Bivalent and Lys4me alone , Figure S1F ) . It is tempting to speculate that the specific localization of Ash1l may be necessary for certain interaction partners of Ash1l , such as Ly4-trimethylases , to be recruited in a promoter-proximal coding region . Of note , we also found that Ash1l was clearly present in the absence of RA ( Figure 3E ) and in genes that were not expressed ( Figures S1F and S1G ) . Surprisingly , it appeared that the methyltransferase activity of Ash1l was inactive under these conditions . Presumably , Ash1l is deposited but poised to achieve an immediate action in response to RA . It remains unclear how the enzymatic activity of Ash1l protein is activated . Future studies on the Ash1l complex and its interaction partners , as well as using knockout mice , may resolve these issues . Unexpected is the increase in the level of Lys36me2 upon DRB treatment . It is possible that , under normal conditions , there may be a competition for methylation sites between Lys36-trimethylase Setd2 and other Lys36-dimethylases including Ash1l . In the presence of DRB , the lack of transcription-dependent trimethylation by Setd2 would result in a spreading of Lys36me2 catalyzed by the dimethylases . In a subset of RAR-associated genes , the Lys36-trimethylase , accompanied with RAR , may generate Lys36me3 on the plat-form of accumulated Lys36me2 in a transcription-independent manner . This may explain the increased levels of Lys36me2/3 upon DRB treatment in the subset of RAR-associated genes including Hoxd4 . In this study , using an Ash1l mutant and DRB , we have revealed a novel function for Ash1l during the establishment of transcriptional activation of Polycomb-regulated genes , including Hox and Wnt family genes . Given that the Wnt signaling pathway integrates numerous environmental signals in vivo , Ash1l may modulate a variety of signals in many biological processes . We have also found novel functional links among several chromatin modifiers that reprogram the status of target chromatin . Future studies on these factors will provide further insights into precise mechanisms for the establishment of transcriptional memory that counteracts Polycomb silencing of developmentally regulated genes .
The animals' care was in accordance with institutional guidelines of National Institute of Genetics in Japan and Saga University Faculty of Medicine . The schematic representation of the strategy used for targeted disruption of mouse Ash1l gene is shown in Figure 1A . A targeting vector was constructed by insertion of DNA fragments of introns 10–12 ( 5′SphI-SpeI ) of mouse Ash1l gene into a ploxFNFDT-SS backbone vector , in which 5′BamHI-3′SphI fragment was replaced to a PCR-cloned floxed exon fragment ( exons 11–12 ) with a Pgk-Neor cassete . PCR primer-pairs used for the cloning are listed in Table S4 . ΔSET mice were generated with M1 ES cells ( derived from F1 of C57BL/6J and 129/Sv ) , and backcrossed to C57BL/6J between two to six times . Genotypes were determined by PCR using the primer-pairs listed in Table S4 . cDNA encoding a part of Ash1l protein ( 2803–2891 , Figure S1C ) was inserted into the bacterial expression vector pGEX 6P-1 ( GE Healthcare ) . The PCR primer-pairs used are listed in Table S4 . GST-fusion proteins were induced and were purified using glutathione-sepharose beads . The eluates containing the recombinant proteins were pooled and dialyzed against PBS . The antibodies were raised against each GST-fusion protein and affinity-purified . Since endogenous Ash1l protein was difficult to detect by immunoblot , the specificity of the antibodies was checked by immunofluorescence analysis under transient expression of lentiviral-mediated shRNA directed against mouse Ash1l mRNA ( Figure S1D ) . Pseudovirus was produced from HEK293T cells by cotransfection of packaging plasmids ( Addgene ) and pRSI9 vector ( Cellecta , Decipher Project ) using PEI-MAX ( Polysciences ) . The target sequence in Ash1l mRNA was following: 5′-GCCAAAUUCUCCUUCUCAUUU-3′ . ES cells were cultured on gelatin-coated dishes in a basic culture medium of KO-DMEM ( Gibco ) containing 1× GlutaMAX-I ( Gibco ) , 1× MEM NEAA ( Gibco ) , 0 . 1 mM 2-mercaptoethanol ( Gibco ) , 50 units/ml penicillin ( Gibco ) , 50 µg/ml streptomycin ( Gibco ) , without feeder cells . For culturing undifferentiated ES cells , the above basic culture medium was supplemented with 1 , 000 units/ml of leukemia inhibitory factor ( LIF ) ( Chemicon ) , 15% Knockout Serum Replacement ( Invitrogen ) , and 1% fetal calf serum ( Gibco ) , and 10 mM 4- ( 2-hydroxyethyl ) -1-piperazineethanesulfonic acid ( Hepes ) buffer . For culturing differentiating ES cells , only 10% fetal calf serum ( Gibco ) was supplemented to the above basic culture medium . A typical protocol for cell culture is shown in Figure 1D , in which RA is added to the differentiation medium at indicated time points . DRB ( Sigma ) was added at a final concentration of 75 µM on either Day 3 or Day 4 ( 16 hour-exposure ) before analysis . Chromatin immunoprecipitation was performed according to online protocols provided by Millipore ( for histone modifications ) or Abcam ( for the other proteins ) with modifications in fixation protocols . The antibodies and fixation protocols used are listed in Table S5 . Immunoprecipitated DNA was purified using a PCR Purification Kit ( Qiagen ) , and was quantified by real-time PCR using SYBR green dye on a LightCycler480 machine ( Roche ) . PCR temperatures for acquisitions of DNA amplification signals were determined empirically . PCR primer-pairs used are listed in Table S4 . Background signals are shown in Figure S11A and are subtracted from most of the respective results . Control ChIP signals in either a promoter-proximal coding region of Gapdh or a promoter region of Il2ra are indicated in relevant figures ( Figure S11B ) . Unless otherwise stated , each result and error bar in graphs represent mean and s . d . , respectively , of three independent PCR reactions from a single ChIP experiment that is representative of several that were performed ( 3 to 5 experiments ) . For ChIP-Seq , 1–5×107 ES cells were used and chromatin was sheared to an average DNA fragment size of 150–250 bp . After immunoprecipitation using Dynabeads protein G ( Invitrogen ) , ChIP-Seq libraries were prepared according to Illumina protocols . The libraries were sequenced using an Illumina HiSeq 1000 . All ChIP-Seq reads were mapped to the mouse genome ( mm9 ) using Bowtie2 with default parameters . Genomic profiles were generated using igvtools and were viewed in Integrative Genomics Viewer ( IGV ) . Peaks of Ash1l and RAR ChIP-Seq signals on genome were determined using MACS2 with false-discovery rate as 0 . 05 . Each associated gene for the peaks was determined using Entrez gene annotation with in-house computer program , in which Ash1l-target genes were defined as genes containing Ash1l-peaks around transcription start site ( TSS ) within +/−4 kb and RAR-associated genes were defined as genes containing RAR-peaks in up-stream ( from −20 kb to TSS ) and coding regions . Datasets for reads/kb/million mapped ( RPKM ) values of Lys36me2/3 in coding regions of each gene were normalized to 75th percentile . Raw sequencing data were submitted to the NCBI Short Read Archive database under accession number ( GSE48421 ) . Mouse ES cell RAR ChIP-Seq datasets ( GSE19409 ) [22] were downloaded from the NCBI Short Read Archive database and were compared with Lys36me2/3 datasets generated by our study . Hox cDNAs were RT-PCR-cloned from embryonic total RNA into pBluescript . Primer-pairs used for PCR amplification are listed in Table S4 . Single-stranded RNA probes labeled with either [35S]-UTP ( for section ) or digoxigenin-UTP ( for whole-mount ) were synthesized according to manufacturer's instructions ( Promega; Roche ) . In situ hybridization was performed according to the procedures described previously [41] , [42] . After hybridization and washing , the sections were immersed in Kodak NBT emulsion ( diluted 1∶1 with 2% glycerol ) , exposed for 2 weeks and developed in a Kodak D-19 developer . For whole-mount in situ hybridization , probes were detected using alkaline phosphatase-conjugated anti-digoxigenin Fab fragment ( Roche ) and signals were developed using Nitro blue tetrazolium chloride ( NBT ) and 5-Bromo-4-chloro-3-indolyl phos- phate , toluidine salt ( BCIP ) ( Roche ) . Skeletal preparations were prepared from perinatal mice as described previously [41] . Cartilage and ossified bone were stained with alcian blue-alizarin red . The run-on transcription assay was performed as described previously with following modifications [43] . Briefly , 5–7×106 cells were treated with ice-cold hypotonic nuclei isolation buffer ( 20 mM Hepes-KOH [pH 7 . 6] , 10 mM NaCl , 5 mM MgCl2 , 0 . 5% NP-40 , 1 mM DTT , 0 . 2 mM PMSF , 1 mM Bezamidine-HCl ) and the isolated nuclei were re-suspended in storage buffer ( 50 mM Hepes-KOH [pH 7 . 6] , 0 . 1 mM EDTA , 5 mM MgCl2 , 40% glycerol ) to give a total 30 µl for each reaction . Transcription was re-started by addition of 30 µl of transcription buffer ( 10 mM Hepes-KOH [pH 7 . 6] , 0 . 3 M KCl , 4 mM DTT ) , 40 units of RNase inhibitor , 3 µl of Biotin RNA Labeling Mix ( Roche ) . The reaction was incubated at 30°C for 45 min on a vortex mixer . After DNase I ( Takara ) treatment , total RNA was isolated using Isogen II ( Nippongene ) and 10–20 µg of total RNA was subjected to further purification of nascent RNA molecules using 50 µl of Dynabeads MyOne Streptavidin T1 ( Invitrogen ) in Click-iT Nascent RNA Capture Kit ( Invitrogen ) . Complementary DNAs were synthesized from purified nascent RNA molecules by on-beads reverse transcription according to the manufacturer's instructions , and the cDNAs were subjected to real-time PCR analyses . Total RNA was prepared using Isogen II ( Nippongene ) and subjected to DNase I ( Takara ) treatment and further purified by aid of RNeasy Mini Kit column ( Qiagen ) . The poly ( A ) -containing mRNA were purified and libraries were prepared according to Illumina TruSeq RNA protocols . Data were obtained with the Illumina HiSeq 1000 sequencing machine . All RNA-Seq reads were mapped to the mouse genome ( mm9 ) using TopHat2 . Transcript abundance was quantified using Cufflinks and annotations from Ensembl release 70 , and FPKM ( fragments/kb of transcript/million fragments mapped ) values were calculated . To minimize dispersion effect by low-FPKM values , all the FPKM values were modified by addition of 0 . 1 in log2 transformation . For a classification of chromatin signature , a supplementary table and ChIP-Seq data in Mikkelsen , et al . [18] were used as references . Gene ontology analysis for biological process of the selected genes was performed using Partek Genomic Suite ( Ryoka systems ) . Raw sequencing data were submitted to the NCBI Short Read Archive database under accession number ( GSE48419 ) . Remaining materials and methods including the method for histone methyltransferase assay are available in Text S1 . | Transcriptional mechanisms in eukaryotes are composed of numerous consecutive steps , including chromatin modification and remodeling . Recent reports using yeast genetics have revealed that Lys36 methylation of histone H3 , a hallmark of the active gene , is a consequence of transcriptional elongation . Similarly , a report using Drosophila genetics showed that exclusion of the Polycomb repressive complexes , general repressor complexes that regulate development and cellular differentiation , is another consequence of transcription . Here , we provide evidence that these causal relationships are not really general . By ceasing ongoing transcription at a certain step using an inhibitor in combination with mouse genetics , we have identified novel intermediate steps of transcription: Ash1l-mediated Lys36 methylation of histone H3 and subsequent exclusion of the Polycomb complexes that occur independently of transcriptional elongation . Furthermore , we show that binding of a nuclear receptor may promote trimethylation of Lys36 in its associated gene independent of its ongoing transcription . In this paper , we detail previously unknown key machineries orchestrated against Polycomb silencing , providing an innovative view of the mechanisms involved in the establishment of transcriptional memory . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Ash1l Methylates Lys36 of Histone H3 Independently of Transcriptional Elongation to Counteract Polycomb Silencing |
The intestine is a common site for invasion by intracellular pathogens , but little is known about how pathogens restructure and exit intestinal cells in vivo . The natural microsporidian parasite N . parisii invades intestinal cells of the nematode C . elegans , progresses through its life cycle , and then exits cells in a transmissible spore form . Here we show that N . parisii causes rearrangements of host actin inside intestinal cells as part of a novel parasite exit strategy . First , we show that N . parisii infection causes ectopic localization of the normally apical-restricted actin to the basolateral side of intestinal cells , where it often forms network-like structures . Soon after this actin relocalization , we find that gaps appear in the terminal web , a conserved cytoskeletal structure that could present a barrier to exit . Reducing actin expression creates terminal web gaps in the absence of infection , suggesting that infection-induced actin relocalization triggers gap formation . We show that terminal web gaps form at a distinct stage of infection , precisely timed to precede spore exit , and that all contagious animals exhibit gaps . Interestingly , we find that while perturbations in actin can create these gaps , actin is not required for infection progression or spore formation , but actin is required for spore exit . Finally , we show that despite large numbers of spores exiting intestinal cells , this exit does not cause cell lysis . These results provide insight into parasite manipulation of the host cytoskeleton and non-lytic escape from intestinal cells in vivo .
Intracellular pathogens have a life cycle that includes three major steps: invasion of the host cell , replication , and exit out of the host cell . While the question of how pathogens invade cells has been intensively studied , much less is known about how pathogens exit cells , although this process appears to be highly regulated [1] . Even when pathogen exit causes lysis of host cells , this lysis is often due not simply to mechanical stress , but is part of a regulated process controlled by the pathogen . For example , the intracellular bacterium Chlamydia uses a cysteine protease to lyse host cells at the proper time as a mechanism of escape [2] . In addition to lytic escape , there are also less destructive modes of pathogen exit . One example of non-lytic pathogen exit comes from the Gram-positive bacterium Listeria , which polymerizes host actin to force its way into neighboring cells , which then engulf the bacterium to allow cell-to-cell spread [3] . Another example of non-lytic pathogen exit involving actin is used by Mycobacterium [4] , [5] . In contrast to Listeria , Mycobacterium appears to break through the host plasma membrane as it is exiting , and the membrane reseals behind the pathogen such that the host cell is not lysed . Acquiring a better understanding of the mechanisms of pathogen exit in vivo could lead to better treatments in a variety of settings , since the process of exit is critical for the propagation and spread of intracellular pathogens of all types . Many intracellular pathogens invade their host and progress through infection in the intestine . However , most studies of pathogen exit have been performed in tissue culture cells or single-celled hosts , such as the studies described above . Unfortunately , these model hosts lack the connectivity , differentiated structures , and polarity of intact intestinal cells . The Caenorhabditis elegans intestinal tract provides an excellent system to study intestinal pathogens as it is composed of 20 epithelial cells that share many morphological properties with human intestinal epithelial cells [6] . In both humans and worms , the intestine contains polarized epithelial cells decorated with apical , finger-like microvilli anchored into a cytoskeletal structure called the terminal web , which is composed of actin and intermediate filaments . Because C . elegans intestinal cells share these structural similarities with human intestinal epithelial cells and because nematodes are transparent , C . elegans provides a very useful whole-animal model for study of host/pathogen interactions in intestinal epithelial cells [7] , [8] . Recently , we described the first natural intracellular pathogen of C . elegans and found that it defines a new species of microsporidia [9] , [10] . Microsporidia are obligate intracellular parasites that can infect virtually all animal phyla , as well as a few protists [11] , [12] , [13] . These parasites comprise a phylum that is either part of the fungal kingdom or is a sister group to the fungi [14] , [15] , [16] , [17] . The Microsporidia phylum contains over 1200 species , 14 of which can cause infection in humans . These infections most commonly afflict AIDS patients and other immunocompromised patients where they can cause persistent and lethal diarrhea [18] . Because so little is known about these microbes and few treatments are available , they have been deemed priority pathogens by the U . S . National Institutes of Health . Microsporidia are also considered microbial contaminants of concern by the U . S . Environmental Protection Agency and can plague agriculturally relevant organisms . For example , microsporidia have been responsible for the collapse of fisheries and have also been implicated in honey bee colony collapse disorder [19] , [20] , [21] , [22] . We named the C . elegans-infecting microsporidia species Nematocida parisii , or nematode-killer from Paris , since it was first isolated from a French wild-caught nematode and it eventually kills its host . The microsporidia are a diverse group of pathogens that can have very complex life cycles . On a general level , the N . parisii life cycle appears similar to that of other microsporidia ( Figure 1A ) [9] , [13] . N . parisii infects the C . elegans intestine in its transmissible spore form and is transmitted horizontally , likely via a fecal-oral route . After ingestion , microsporidia invade host cells using a specialized infection apparatus called a polar tube , which is coiled inside of the spore and then “fires” to pierce the host cell . The polar tube can inject the nuclei and sporoplasm of the parasite directly into host cells , thereby avoiding extracellular defenses of the host . This injected parasite material develops intracellularly in a replicating cellular structure called a meront . These meronts go through several stages of development and eventually re-differentiate to generate mature spores that somehow must exit the cell and continue the parasite life cycle . N . parisii ultimately kills its host , but C . elegans can sustain a large spore burden before death . Indeed , live animals can be contagious to their neighbors , indicating that there is a mechanism of exit that does not cause severe damage to the host . Previously , we had observed gaps in the intermediate filaments of the terminal web in infected animals and hypothesized that they may be part of a regulated exit strategy of N . parisii [9] . However , we did not know what initiated these gaps , nor how they related to pathogen development and exit . Here we show that N . parisii infection causes dramatic rearrangement of intestinal actin that is part of a novel two-phase , non-lytic exit strategy . The first phase involves restructuring of the terminal web , which is a barrier that pathogens must cross to exit cells . During this phase the subcellular localization of actin is altered , with the apically-restricted actin appearing ectopically on the basolateral side of host cells and forming networks . Subsequently , gaps appear in the terminal web , as assessed by IFB-2 , an intermediate filament component of the terminal web . We hypothesize that actin redistribution away from the apical membrane may trigger these gaps in the terminal web , since lowering actin levels causes terminal web gaps in the absence of infection . Soon after the appearance of gaps , the parasite enters the second phase of the exit strategy . During this phase N . parisii spores are able to exit the cell . Interestingly , we find that host actin promotes spore exit and that decreasing the levels of actin impairs spore exit . Our analyses of cell integrity indicate that despite very large numbers of spores exiting C . elegans intestinal cells , these cells do not lyse . We propose that a novel , two-phase , non-lytic exit strategy allows N . parisii to escape from host intestinal cells while minimizing damage to its host , thus increasing parasite transmission .
N . parisii infection proceeds through distinct stages in C . elegans intestinal cells , starting with the meront stage when the parasite is actively replicating ( Figure 1A ) . N . parisii meronts appear to develop in direct contact with the host cytoplasm , essentially creating “parasite organelles” , which eventually develop into mature spores [9] . Both meronts and spores form inside C . elegans intestinal cells , which are polarized epithelial cells that are found in rings of two to four cells that form a tube [6] . Like mammalian intestinal cells , these cells are decorated on the apical ( lumenal ) side with microvilli that are anchored into a cytoskeletal structure called the terminal web . The C . elegans terminal web contains a specialized actin isoform called ACT-5 and an intermediate filament component called IFB-2 ( Figure 1B , C ) . ACT-5 is restricted mostly to microvilli-containing cells and is localized to both the microvilli and the terminal web [23] , whereas IFB-2 is restricted to the terminal web [24] , [25] . To examine how N . parisii infection modifies the intestinal cytoskeleton , we generated a strain that expresses YFP::ACT-5 as well as IFB-2::CFP in the intestine to track both of these cytoskeletal components simultaneously . In uninfected animals , we found that YFP::ACT-5 colocalized with IFB-2::CFP at the apical side of intestinal cells ( Figure 2A ) , consistent with previous reports [24] , [25] . We next imaged these cytoskeletal components during infection with N . parisii . Because N . parisii cannot be propagated outside of host cells , we prepare infectious spores by mechanically disrupting infected animals and purifying the spores away from other material . By adding a preparation of these spores onto plates regularly used to maintain C . elegans , the animals become infected by simply ingesting the spores . Once ingested , the spores then presumably fire their polar tubes , injecting N . parisii sporoplasm and nuclei into the C . elegans intestinal cells . We investigated cytoskeletal protein localization as N . parisii infection progressed and found that infection with N . parisii caused dramatic changes in ACT-5 localization without concomitant changes in IFB-2 . The first relocalization of ACT-5 occurred during N . parisii meront development , which is the first stage of parasite development . At this point , we found that ACT-5 localization no longer remained restricted to the apical side of intestinal cells , but appeared ectopically on the basolateral side of cells ( Figure 2B ) . Coinciding with this ectopic basolateral localization that appeared as an unbranched line , we also observed more complex networks of ACT-5 expression ( Figure 2C ) . These networks exhibited branches that varied in number and length , and also appeared basolaterally . The exact molecular nature of these structures is unknown , so we refer to them generally as “actin networks” . They may be part of the same basolateral relocalization phenomenon shown in Figure 2B , but are more easily visualized in the plane of focus shown in Figure 2D because of the orientation of the animal on the slide . In both cases , actin still remained apically localized , although often at lower levels . Videos S1 and S2 show a Z-series of YFP::ACT-5 localization in the intestine to provide a three-dimensional view of this actin relocalization phenomenon . In general , we did not observe IFB-2 relocalizing with ACT-5 ( Figure 2B and 2C ) , indicating that N . parisii infection directs a specific relocalization of actin . To determine whether these ectopic patterns of ACT-5 localization were general responses to pathogen infection or were specific to N . parisii infection , we analyzed YFP::ACT-5 localization in animals infected with other intestinal pathogens . The Gram-negative bacterial pathogens Pseudomonas aeruginosa and Salmonella enterica also cause lethal intestinal infections in C . elegans [26] , [27] , [28] , [29] . We did not observe ectopic actin basolateral localization or networks in animals infected with P . aeruginosa ( n = 55 animals ) , or S . enterica ( n = 154 animals ) , indicating that the previously described localization patterns are somewhat specific to infection by the natural intracellular parasite N . parisii . We also examined a larger number of uninfected animals and saw only one out of 200 animals that had a single ectopic actin structure without obvious branching , further supporting our finding that this dramatic actin relocalization is specific to N . parisii infection . To further characterize this actin relocalization in infected animals , we examined the kinetics of ACT-5 ectopic localization with respect to changes in IFB-2 and to progression of N . parisii development . We infected a synchronized population of YFP::ACT-5;IFB-2::CFP animals and tracked both N . parisii parasite development and changes in these cytoskeletal markers in a population over time ( Figure 3 ) . The stages of parasite development are shown in Figure 3A–D . We subdivided the meront stage into “early meront” ( Figure 3B ) , where there was only a small area of gut granule clearing in the intestine , and “late meront” ( Figure 3C ) , where there was more extensive clearing throughout the intestine . Following meront development , spores become visible ( Figure 3D ) . Early meronts appeared by 16 hours post-inoculation ( hpi ) , late meronts appeared by 30 hpi and spores appeared by 40 hpi ( Figure 3E ) . The first obvious change in the cytoskeleton was the ectopic localization of YFP::ACT-5 ( Figure 3E ) , appearing as a line on the basolateral side of intestinal cells ( as shown in Figure 2B ) . This pattern was observed in all animals of a population during the late meront stage , at 30 hpi . In addition to this linear basolateral relocalization , there was also evidence of ACT-5 networks at this time ( Figure 2C and Video S2 ) , continuing at a low level until 40 hpi when all animals contained spores . We also examined how this ectopic actin localization correlates with restructuring of the terminal web by assessing IFB-2 localization in these same animals . Previously , we had identified gaps in the terminal web of infected animals using transmission electron microscopy ( TEM ) and gaps in IFB-2 localization using anti-IFB-2 antibodies [9] . Here , we found very similar-appearing gaps in infected animals using CFP-tagged IFB-2 ( IFB-2::CFP ) ( compare uninfected animals in Figure 3F to heavily infected animals in Figure 3G ) . We first observed these gaps in 13% of animals infected at 30 hpi , the late meront stage of infection ( Figure 3E ) . At 40 hpi , when all animals were infected at the spore stage , gaps were found in the terminal web of 100% of animals ( Figure 3E ) . Based on these observations , it appears that actin relocalization occurs before IFB-2 restructuring during infection . We saw a similar pattern of IFB-2 restructuring with respect to parasite development in single IFB-2::CFP transgenic animals , indicating that the presence of the YFP::ACT-5 transgene did not affect IFB-2::CFP expression or localization ( Figure S1 ) . The studies described above were performed with a population of animals , which is informative , but does not provide insight into the progression of symptoms within an individual animal . To determine whether the trends we observed in a population ( e . g . ACT-5 is relocalized before IFB-2 localization is disrupted ) were also true in an individual , we performed longitudinal studies . Individual animals were mounted on slides at varying timepoints for viewing , and then recovered onto plates in between these timepoints to allow them to recover and to allow the infection to progress ( Table S1 ) . Consistent with the population studies , these experiments showed that YFP::ACT-5 basolateral relocalization occurred first , followed by gaps in IFB-2::CFP localization . Our previous studies with IFB-2 antibodies had suggested that terminal web gaps were specific to N . parisii infection – we did not observe them in animals infected with other pathogens that cause lethal intestinal infections , such as P . aeruginosa and Staphylococcus aureus [9] . Using the IFB-2::CFP transgene we also analyzed the terminal web in animals infected with the bacterial pathogen S . enterica , but did not find evidence that infection caused gap formation in the terminal web ( n = 154 animals ) . To further confirm that terminal web gaps were a N . parisii-induced phenotype , we used the IFB-2::CFP transgenic animals and performed a thorough inspection of IFB-2 localization in uninfected animals ( n = 200 animals ) . We found that only one animal exhibited a gap , and this was a single , isolated gap . For our analyses of N . parisii infected animals , we only scored animals as positive for gap formation if we observed multiple gaps . Thus , similar to the ACT-5 relocalization phenotype , the gaps observed in IFB-2::CFP localization appear to be specific to N . parisii infection . Since relocalization of ACT-5 during N . parisii infection preceded the appearance of gaps in IFB-2 localization , we hypothesized that reduced levels of ACT-5 in the apical membrane might cause IFB-2 terminal web gaps . To test this hypothesis we reduced ACT-5 expression with feeding RNA interference ( RNAi ) against act-5 . We observed a substantial reduction in YFP::ACT-5 expression in act-5 RNAi-treated animals compared to control-treated animals , confirming the efficacy of act-5 RNAi ( Figure 4A , B ) . Interestingly , when we analyzed the terminal web in act-5 RNAi-treated animals we found gaps in IFB-2::CFP that appeared in the absence of infection ( Figure 4B ) . Therefore , we speculate that infection-induced reduction of ACT-5 at the apical membrane may be a triggering event that results in the disruption of the terminal web . Our studies of N . parisii parasite development and host cell restructuring suggested that terminal web gap formation is an important stage of the infection cycle . Therefore , we investigated whether terminal web gaps have a functional role by examining how they relate to spore exit . As a functional read-out for spore exit , we tested the contagiousness of individual infected animals: in order for an animal to be contagious , spores must exit intestinal cells , leave the intestinal tract ( likely through normal defecation ) , and then be consumed by a neighboring animal . To measure contagiousness , uninfected “recipient” animals were exposed to individual infected “donor” animals for a defined length of time . Each individual donor animal was then removed and examined for stage of infection and whether it had gaps in IFB-2::CFP localization . Several days later , recipient animals were examined to determine whether they had become infected , thus implying that the donor animal had shed spores and was contagious . After testing 164 individual donor animals for contagiousness , we found 87 animals that were contagious and all of these animals had spores , consistent with previous studies . Strikingly , we found that 100% of these contagious animals exhibited terminal web gaps ( Figure 5A ) , consistent with the model that gaps are functionally related to spore exit . Of the animals that were non-contagious , 12% of them exhibited a small number of spores ( and gaps ) . In previous studies we had found that animals infected with only a small number of spores are not always contagious , likely due to a low amount of spores being shed before the donors were removed from the plate [9] . There were also 9% of non-contagious animals that exhibited gaps but only meronts ( no spores ) , consistent with population studies that indicated gaps appear slightly before spore formation ( Figure 3E ) . The remaining 79% of animals that were non-contagious exhibited no gaps and only meronts ( no spores ) . Since all contagious animals exhibit terminal web gaps , these data support the model that terminal web gaps may be a necessary part of spore exit from the intestinal cells . We next sought to determine whether new gaps are continually being made during the course of infection , or whether gap formation is a discrete event that occurs only once during infection . In general , gaps first appeared in the region of the intestine exhibiting the heaviest N . parisii infection and gaps were always present in the region of intestine that had spores ( Table S1 and data not shown ) . To quantify the number of gaps over time we chose a plane of focus with the largest number of gaps from N . parisii spore-infected animals and counted the number of gaps at 40 hpi , when they first are visible in all animals , and again at 63 hpi . With this analysis , we found that the number of gaps per unit area remained similar as the infection proceeded , suggesting that new gaps are not made after their initial formation at 40 hpi ( Figure 5B ) . Next , we quantified the size of gaps at 40 hpi and 63 hpi ( see Figure legend and Experimental Procedures for more detail on quantification ) . With this analysis , we found that gaps became larger as the infection proceeded , suggesting that once a gap is made it becomes larger over time ( Figure 5C ) . This effect may be due to the lumenal distention that is observed as infection progresses , or it may be due to a more active process . In any case , these data indicate that the formation of gaps appears to be a discrete , regulated event that is orchestrated by the parasite at a particular stage in infection . The extensive relocalization of ACT-5 during infection and the induction of gaps in IFB-2 localization upon RNAi against act-5 led us to investigate the functional role of ACT-5 during infection . First , we assessed whether reducing levels of actin would affect the development of N . parisii infection inside intestinal cells . We measured the progression of N . parisii infection in animals fed with bacteria expressing dsRNA against act-5 . Because feeding undiluted act-5 dsRNA-expressing bacteria can slow the growth of animals , we titrated back the dose of RNAi to allow animals to develop more normally . Diluting act-5 RNAi bacteria 1:25 and 1:50 with control bacteria allowed worms to develop relatively normally , but still caused reduced act-5 levels ( as assessed by imaging YFP::ACT-5 , Figure S2 ) . We found that meronts and spores appeared in act-5 RNAi-treated animals at a similar rate to control-treated animals ( Figure 6A ) . Therefore , despite the changes induced in actin localization upon infection , reduction of act-5 expression had no obvious effect on progression of N . parisii infection or formation of spores within the host cells . Since reducing actin in the absence of infection caused gaps in the terminal web , which could be a barrier for exit , we speculated that reducing actin might cause animals to shed more spores . Thus , we sought to measure spore exit in act-5-defective animals . Previously , we used single animal contagiousness assays to measure spore exit ( Figure 5A ) . However , these assays are labor-intensive and categorical ( animals are binned as either contagious or non-contagious ) . Therefore , we developed a more quantitative “spore shedding” assay to measure the number of spores shed by a population of animals . Fifty infected animals were placed in liquid media containing E . coli ( the normal food source of C . elegans ) and allowed to shed spores into the media . After defined periods of time , the number of spores shed was quantified . Animals appeared to shed spores throughout the assay , as we found increasing numbers of spores were shed with increasing amounts of time ( Figure 6B ) . Under these assay conditions we found that a single infected worm could produce thousands of spores per hour . However , the number of spores shed per animal had significant animal-to-animal variability ( data not shown ) , so we performed assays with populations of animals . Using this assay , we tested whether reducing ACT-5 levels would increase N . parisii spore shedding . Surprisingly , we found that spore shedding was almost completely blocked in 1∶25 act-5 RNAi-treated animals and was substantially blocked in 1∶50 act-5 RNAi-treated animals compared to control-treated ( L4440 ) animals ( Figure 6C ) . These findings suggest that reduction of ACT-5 impairs spore exit , in contrast to our original hypothesis . Altogether , our experiments indicate that proper levels of host actin are not important for N . parisii development inside cells , but are critical for promoting spore exit from those cells . To further test the hypothesis that actin promotes spore exit we examined two other C . elegans strains that have differing levels of act-5 expression . First , we examined the YFP::ACT-5 transgene , which likely overexpresses actin . Our experiments indicated that YFP::ACT-5 had no effect on the levels of spore shedding in our assays compared to wild-type animals ( Figure 6D ) . Next we examined the act-5 ( ok1397 ) deletion mutant , which is likely a null mutation , as it deletes half of the 5′ translated region of act-5 as well as the upstream untranslated region . Because the act-5 ( ok1397 ) mutation is lethal when homozygous , we examined spore shedding in act-5 ( ok1397 ) /+ heterozygote mutants , which develop normally . We infected these animals with N . parisii and found that spore development proceeded at a similar pace as in wild-type animals ( data not shown ) . Interestingly , in act-5 heterozygote mutants where act-5 expression should be reduced to half of wild-type levels , we found that spore shedding was approximately half of wild-type levels ( Figure 6D ) , consistent with our hypothesis that actin promotes exit . This result suggests that there is a dose-dependent effect of actin on spore shedding , and that proper levels of actin are critical for efficient N . parisii spore exit from host cells . The spore shedding and contagiousness assays described above measure the spores that have left the animal , presumably through defecation . Before these spores are shed through defecation , they must first exit the intestinal cells into the lumen . Our previous studies indicated that spores are only found in intestinal cells , and spores are therefore unlikely to exit out of cells basolaterally to enter the rest of the animal ( e . g . into the gonad ) . To more closely examine that spores only exit apically as opposed to basolaterally , we performed transmission electron microscopy ( TEM ) to analyze the location of spores in the whole animal . The only tissue in which we conclusively found spores was the intestine ( Figure 7A ) ; we did not see evidence of spores in any other tissue ( e . g . hypodermis , muscle or gonad ) . In order to further examine spore location throughout the animal , we also fixed animals and stained with Calcofluor White , a dye that binds to the chitin found in N . parisii spores . Again , we only found evidence of N . parisii spores in the intestine , not in other regions of the animal ( Figure 7B ) . Because N . parisii spores have to cross the C . elegans apical plasma membrane of intestinal cells to exit into the lumen , we investigated whether spores in the lumen may have acquired C . elegans plasma membrane upon exit . To address this question we infected transgenic animals expressing PGP-1::GFP , which is a GFP-tagged ATP binding-cassette ( ABC ) transporter that localizes to the apical side of intestinal cells [30] . After infecting worms , we transferred them repeatedly away from spores onto fresh plates , so that any spores in the lumen were likely the result of exit and not the result of consumption ( see Materials and Methods for details ) . We did not find any lumenal N . parisii spores that were labeled with GFP ( Figure 7C , n = 50 spores examined ) , suggesting that spores do not take C . elegans membrane with them when they exit into the lumen . It is clear that N . parisii spores are exiting while animals are still alive , as live animals are contagious to other animals ( [9] and Figure 5A ) and animals are alive at the end of spore shedding assays ( Figure 6 ) . However , it is unclear whether this process of spore exit causes tissue damage that leads to cell lysis . To address this question , we next examined whether N . parisii spore exit and/or terminal web restructuring cause disruption of cellular integrity of the intestinal cells . To assay cellular integrity , we used the cell-impermeable dye propidium iodide ( PI ) , which can be fed to C . elegans . When intestinal cells are intact , PI does not enter cells after short incubation periods , but instead remains restricted to the intestinal lumen . As a positive control for PI entry into cells that have lost integrity , we used animals that were exposed to the Bacillus thuringiensis pore-forming toxin Cry5B , which creates nanometer-sized holes in the plasma membrane . A recent study demonstrated that Cry5B treatment of C . elegans creates pores in intestinal cells that allow PI to enter intracellularly [31] . Consistent with these studies , we found that animals fed E . coli that expresses Cry5B had intracellular PI staining whereas animals fed normal E . coli did not ( compare Figure 8A and B ) . Strikingly , animals infected with N . parisii ( also fed on E . coli ) did not exhibit intracellular PI , indicating that cellular integrity was not compromised ( Figure 8C and G ) . These animals were infected with N . parisii spores ( Figure 8H ) , and exhibited gaps in IFB-2::CFP localization ( Figure 8I ) . In order to quantify this effect , we scored the percentage of animals with intracellular PI in 100 animals of each condition , only choosing animals that were infected throughout the intestine with spores for the N . parisii infection condition . From single-animal contagiousness assays with animals at a similar stage of infection ( Figure 5A ) , we estimate that 80-90% of these animals were contagious . In these studies we found that almost all Cry5B-treated animals exhibited intracellular PI , whereas virtually none of the N . parisii-infected animals did ( Figure 8J ) . Thus , while N . parisii infection causes gaps in the terminal web and results in the passage of 1 µm-wide N . parisii spores ( much larger than Cry5B-induced pores ) across the membrane in order to exit into the lumen , these events do not appear to cause lysis of C . elegans intestinal cells . These data support the hypothesis that cytoskeletal rearrangements and spore exit are highly regulated processes that maintain the integrity of the host cell during infection .
Intracellular pathogens have privileged access to host cell components that they can exploit in order to survive , replicate , and then exit to infect new hosts . The studies presented here provide insight into parasite-directed cellular reorganization that occurs in vivo in order to exit cells but minimize damage to the host . We find that the intestinal parasite N . parisii redirects host actin in order to non-lytically exit from intestinal cells and propose a two-phase model for this exit , with terminal web restructuring occurring during Phase I and non-lytic exit occurring during Phase II ( Figure 8K ) . Phase I takes place after invasion ( step 1 ) , when N . parisii is developing as a metabolically active meront , a stage that essentially forms a new organelle within the host cell . At this time , the intestinal-specific actin isoform ACT-5 ectopically relocalizes to the basolateral side of intestinal cells , often forming network-like structures ( step 2 ) . Soon after this relocalization of ACT-5 , gaps appear in localization of IFB-2 , an intermediate filament component of the terminal web ( step 3 ) . We speculate that relocalization of ACT-5 may trigger these gaps , since reducing the level of ACT-5 expression causes gaps in the terminal web in the absence of infection . We show that all contagious animals exhibit gaps in localization of IFB-2 , and that formation of gaps occurs during a discrete stage of the infection cycle , suggesting that gap formation is a highly regulated process . Phase II of the exit strategy begins after N . parisii develops into spores and these spores exit out of cells into the lumenal space ( step 4 ) . Surprisingly , we found that reducing levels of ACT-5 impairs spore shedding , indicating that host actin is required for proper exit of N . parisii spores at this stage , even though the infection and parasite development can proceed normally within the cell . Interestingly , N . parisii spore exit is non-lytic , since extracellular dyes continue to be excluded from the intracellular space of intestinal cells ( step 5 ) . This finding is striking , as a single infected animal can shed thousands of spores ( Figure 6 ) . Our functional analyses of host intestinal actin during N . parisii infection indicate that there are two distinct roles for actin , with opposing effects in each of the two phases of exit . As described above , it appears that host actin impairs exit during Phase I , since a reduction of actin leads to gaps in the terminal web , which may present a barrier to exit . However , we found that a reduction of actin levels led to an overall reduction in spore exit . Thus , one explanation is that host actin impairs spore exit during Phase I , but promotes spore exit during Phase II . A caveat to this interpretation is that the function of terminal web gaps ( and thus the role of actin in Phase I ) is unclear since we do not yet have a method to block gap formation during infection: it is possible that gaps serve a purpose other than facilitating exit . Our proposed role of actin in Phase II to promote exit also requires further investigation . One possibility is that actin provides motor force to drive spores out of the cell . The nature of this actin polymerization and its relocalization at the proper time may be the key to orchestrating the complex events that enable large numbers of microsporidian spores to egress out of C . elegans intestinal cells in a manner that minimizes damage to its host . Understanding how N . parisii is able to exit the host cell necessitates understanding what barriers it must cross and the mechanism it employs to cross them . In Phase I of the N . parisii exit strategy , the terminal web is restructured . Our studies indicate that terminal web gaps appear right before spore formation ( Figures 3E and 5A ) , implying that gaps are not caused by the mechanical process of spore exit , but rather by a regulated signaling event that is precisely timed to precede spore formation and the need to exit . The terminal web is a conserved structure found in microvilli-containing cells from C . elegans to humans , but surprisingly little is known about its assembly and dynamics [32] , [33] , [34] , [35] . The terminal web presumably represents a barrier that host vesicles must traffic through during normal endocytosis and exocytosis events . Do intracellular vesicles need to dissolve the terminal web in order to cross this barrier ? Electron microscopy studies indicate there are tiny gaps in the C . elegans terminal web that are too small to allow for vesicle passage , but are associated with vesicular elements and could represent a system of regulated passage [36] . Perhaps N . parisii is exploiting this vesicle passage system to create gaps in the terminal web , but in a way that does not result in resealing of these gaps . It will be interesting to perform live imaging of the kinetics of these structures in uninfected animals to examine what is the basal level of movement and restructuring of this conserved structure in intestinal cells . It will also be interesting to further examine the mechanism by which N . parisii spores cross the plasma membrane , as our studies indicate that they do not acquire C . elegans membrane as they exit into the lumenal space of the intestine ( Figure 7C ) . In spite of the terminal web restructuring described above , and the passage of 1 µm-wide spores out of cells into the digestive tract , cellular integrity assays indicate that actin-based exit of N . parisii spores does not cause lysis of C . elegans intestinal cells ( Figure 8 ) . This finding is intriguing , since bacterial pathogens such as Listeria and Mycobacterium also use actin-based , non-lytic forms of exit , suggesting a common evolutionary strategy for exit between eukaryotic and prokaryotic pathogens . Several bacterial virulence factors have been identified that direct these actin-based processes within host cells [4] , [37] , but comparatively little has been explored about eukaryotic intracellular pathogens and the factors they use to exploit host cells . And almost nothing is known about virulence factors in the Microsporidia phylum . One of the few characterized examples of microsporidian virulence factors is a class of ATP transporters that can “steal” ATP from the host cell [14] , [38] . These transporters are expressed on the plasma membrane of the parasite while it is living inside the host cell and act to import ATP directly into the parasite . In addition to this method of host exploitation , it is likely that microsporidia also secrete factors from the meront into the host cytoplasm to facilitate other kinds of nutrient acquisition , as well as to direct the cytoskeletal changes we have observed . The N . parisii genome has recently been sequenced ( E . R . T . and the Broad Institute , as part of the Microsporidian Genomes Consortium , unpublished data , http://www . broadinstitute . org/files/shared/genomebio/Microsporidia_wp . pdf ) and may provide clues into which parasite components are secreted into the host cell to direct this restructuring . We have isolated microsporidia-infected nematodes from a wide variety of geographical locations , including multiple regions in France ( the N . parisii strain used in this study was isolated near Paris ) , Portugal , India , Colombia and Cape Verde ( [9] and Marie-Anne Félix , personal communication ) . Thus , Caenorhabditis nematodes have likely co-evolved with Nematocida and other microsporidia species . Over time , co-evolution of host/parasite pairs is thought to lead to a reduction in virulence [39] , [40] . This pressure is likely to be especially great in obligate intracellular parasites such as microsporidia , which cannot grow in the absence of host cells . Microsporidian species that infect fish appear to have adopted a strategy of minimal virulence in order to maximize parasite production: they often form “spore factories” called xenomas in the fish nervous system that can produce large numbers of spores without substantial impact on the health of the fish [20] . Perhaps the non-lytic exit of N . parisii is part of a strategy similar to the xenomas found in fish that serves to maximize spore production and transmission , but minimize virulence . N . parisii likely has an intimate relationship with the C . elegans intestinal cell during the meront stage , since it essentially creates a parasite organelle that develops in direct contact with the host cytoplasm [9] . The human-infecting microsporidian pathogen Enterocytozoon bieneusi also develops in direct contact with the host cytoplasm during the meront stage [41] . E . bieneusi infection in humans is restricted to intestinal cells , and is the most common cause of microsporidian disease in humans , being responsible for lethal diarrhea in AIDS patients [18] . Effective treatments are lacking for E . bieneusi infection [15] , [42] and it has not been possible to propagate E . bieneusi in tissue culture cells , perhaps because these cells lack some aspect of differentiated intestinal cells that is needed for the infection cycle . While E . bieneusi and N . parisii are in distinct clades of the microsporidia , their similar developmental life cycle in the cytoplasm of intestinal cells may involve similar pathogenic strategies . The C . elegans/N . parisii host/parasite system may thus provide insights into the mechanisms employed by medically relevant but intractable microsporidian species such as E . bieneusi to further understand how they cause disease and potentially how to treat the infections they cause .
C . elegans were maintained on NGM plates seeded with OP50-1 , as described [43] . We used N2 wild-type animals . BJ49 kcIs6[IFB-2::CFP] was kindly provided by Olaf Bossinger . JM125 caIs[ges-1p::YFP::ACT-5] was kindly provided by James McGhee [24] . ERT38 caIs [YFP::ACT-5;IFB-2::CFP] was made by crossing BJ49 and JM125 strains . A separately made YFP::ACT-5 strain , IN4000 dtIs2298 , was kindly provided by James Waddle and used for Videos S1 and S2 . VC971 +/mT1 II; act-5 ( ok1397 ) /mT1[dpy-10 ( e128 ) ] III was provided by the C . elegans Reverse Genetics Core Facility at UBC via the Caenorhabditis Genetics Center ( CGC ) . The presence of the act-5 ( ok1397 ) deletion was confirmed by PCR genotyping . GK288 GFP::PGP-1 and E . coli ( OP50 ) -Cry5B strains were kindly provided by Ferdinand Los and Raffi Aroian . N . parisii strain ERTm1 ( isolated originally from Franconville , France , near Paris ) was used for all microsporidia infection experiments . N . parisii infected animals were disrupted with silicon beads as described [9] . This lysate was then filtered through a 5 µm filter ( Millipore ) attached to a 10 ml syringe , to eliminate undisrupted C . elegans eggs , larvae and other debris . The filtrate containing N . parisii spores was frozen at −80°C and then thawed right before use . N . parisii spores were quantitated by staining with Calcofluor White ( Fluka ) and counting with a hemocytometer ( Cell-Vu ) . In general , synchronized L1 larvae were grown for 24 hours on 10 cm NGM plates seeded with OP50 at 20°C until approximately L3 stage , when they were infected with 3×107 N . parisii spores and then incubated at 25°C . Symptoms of infection were tracked by mounting approximately 50 animals on agarose pads and then viewing animals with Nomarski optics and/or fluorescence on a Zeiss AxioImager at 630X magnification . Images were captured with Axiovision software . P . aeruginosa PA14 infection assays and S . enterica SL1344 infection assays were performed as described [44] , except that S . enterica infection assays included a pre-treatment with bec-1 RNAi to knock down the autophagy pathway , as described [29] . Transmission electron microscopy was performed as described [9] . Calcofluor White staining on infected animals was performed by fixing and staining with a 1∶1 mix of 1M NaOH and Calcofluor White . In order to examine spores that had newly exited into the intestinal lumen ( as shown in Figure 7C ) , animals were repeatedly washed to reduce exposure to external spores they might ingest . Specifically , at 24 hpi ( prior to spore formation ) animals were washed four times in M9 , added to fresh NGM plates seeded with OP50 , and incubated at 25°C until 40 hpi when this process was repeated . To reduce the probability of animals consuming spores shed by their neighbors , worms were replated at a low density of approximately 45 worms per 6 cm plate . Lumen-localized spores were then imaged in animals at 42 hpi . Donor BJ49 [IFB-2::CFP] animals were generated for contagiousness assays as follows . Synchronized L1 larvae were grown on 6 cm NGM plates seeded with OP50 for 24 hours at 20°C , then infected with 1 . 5×105 , 4 . 7×105 or 9 . 3×105 spores diluted in 500 µl of M9 and placed at 20°C or 25°C for 20 hours in order to generate animals at a variety of infection stages . These animals were picked to new NGM/OP50 plates without spores and incubated for 4 hours , washed off the plate with M9 , rinsed three times to remove spores attached to the cuticle and then added to a new plate . Next , these infected donor animals were individually placed onto 6 cm NGM plates seeded with OP50 as well as 200 L1 recipient animals . The donor and recipient animals were co-incubated for 16 hours and the donor animal was then removed to assess infection level by mounting the animal on a slide and scoring for meronts or spores using Nomarski optics , as well as scored for gaps using fluorescence microscopy , at 630X . Recipient animals were scored for infection similarly 7 days later . If no infection was observed , plates were kept and then scored again for infection 3 days later . Animals at 40 and 63 hours post-inoculation ( hpi ) were analyzed . A 10 µm length of the intestine was chosen such that the wall of the intestine was in a single plane of view . The width of this 10 µm long section was measured and the total number of gaps in IFB-2::CFP localization was counted to calculate the number of gaps per µm2 . This area was then divided into 4 quadrants . In a single , randomly chosen quadrant , the area of each individual gap within the quadrant was quantified using the measure tool in AxioImager in order to calculate the average size of gaps . Statistical comparisons of data at 40 hpi and 63 hpi were done with a two-tailed t-test in Excel . ERT38 [IFB-2::CFP; YFP::ACT-5] synchronized L1 animals were plated on NGM OP50 plates and infected with 3 . 6×106 spores 24 hours later . In order to perform longitudinal analysis of infection symptoms in the same animal over time , animals were anesthetized with levamisole and mounted onto agarose slides for scoring of symptoms , and then recovered onto NGM/OP50 plates until the next timepoint . Feeding RNAi was performed as described . Briefly , RNAi bacterial clones were streaked onto LB agar plates with ampicillin and tetracycline . A single colony was inoculated into LB with ampicillin , grown overnight and then seeded on RNAi plates ( essentially nematode growth media with IPTG and ampicillin – usually 6 cm plates ) . Plates were incubated for at least one day before being seeded with a synchronized population of L1 C . elegans . act-5 RNAi was either used undiluted , or diluted 1∶25 or 1∶50 with L4440 ( vector alone ) control RNAi bacteria . We performed act-5 RNAi experiments either with a clone from the Ahringer RNAi library or a clone that was a gift from James Waddle [23] . RNAi clones were confirmed by sequencing . Similar results were obtained with both clones , but in general , the Ahringer clone caused more severe phenotypes . Animals were infected with 4 . 5×105 or 9 . 3×105 N . parisii spores on NGM plates seeded with E . coli OP50 and placed at 25°C . At 24 hpi , animals were washed with M9 and transferred to new NGM OP50 plates ( except when performing RNAi ) . Then at 48 hpi animals were washed several times and then transferred to a fresh NGM/OP50 plate to remove external spores . Next , 50 animals were picked off of this plate into a microfuge tube with M9 and concentrated OP50 , and rotated on a nutator for 16 hours . The microfuge tube was removed from the nutator and incubated without shaking for 10 minutes to allow the C . elegans , but not the spores , to pellet to the bottom of the tube . Then the supernatant was transferred to a new microfuge tube and the spores were pelleted by spinning at high speed in a microfuge . The supernatant was removed after this spin and the spores in the pellet were stained with Calcofluor White and quantified with a hemocytometer . The absolute number of spores shed per worm varied from experiment to experiment . The viability of animals was confirmed at the end of the assays and assays were performed in triplicate for each experiment . Statistical comparisons of data were done with a two-tailed t-test in Excel . Assays for integrity of intestinal cells was performed with a propidium iodide assay , as described [31] . Briefly , YFP::ACT-5; IFB-2::CFP animals were infected with 9 . 3×105 spores on an OP50-seeded 6cm NGM plate . Approximately 40 hours later , animals were washed off the plate and put in a solution of 5-HT at 5 mg/ml for 15 minutes in order to force feed dye into animals – Cry5B treatment causes animals to cease feeding . Animals were then stained in a 20 µg/ml solution of propidium iodide for 30 minutes , washed twice with M9 to remove background excess dye , and imaged using fluorescence microscopy at 630X . We scored only animals that were at the “full spore” stage and therefore were very likely contagious and actively shedding spores . Negative control animals were treated similarly except they were not infected with spores . Positive control Cry5B-treated animals were plated as synchronized L1 larvae on NGM OP50 plates until they reached the L4 stage . They were then washed onto NGM/Ampicillin/IPTG plates seeded with OP50-Cry5B and incubated at 25°C for 30 minutes . They were then washed off the plates and treated as detailed above . In order to quantify the percentage of animals exhibiting propidium iodide staining intracellularly , we imaged animals with a defined exposure time and then measured the maximum fluorescence intensity in the lumen and intracellularly using ImageJ software from the NIH ( http://rsb . info . nih . gov/ij/ ) . If the maximum fluorescence intensity was greater intracellularly than in the lumen , the animal was binned as exhibiting intracellular propidium iodide . Statistical comparisons of data were done with a two-tailed t-test in Excel . Accession numbers for the genes and gene products mentioned in this paper are given for Wormbase , a publically available database that can be accessed at http://www . wormbase . org . These accession numbers are: act-5 ( T25C8 . 2 ) , ifb-2 ( F10C1 . 7 ) , bec-1 ( T19E7 . 3 ) , pgp-1 ( K08E7 . 9 ) . | The intestine is a common site for invasion by pathogens , but little is known about how pathogens exit out of live intestinal cells in order to spread and propagate . One group of parasites that often invades the intestine is the microsporidia , which comprise a phylum of over 1200 fungal-like species that can cause disease in humans , as well as in agriculturally significant organisms such as fish , silkworm and honey bee . Here , we investigated a natural microsporidian infection in live intestinal cells of the roundworm C . elegans . We discovered a novel exit strategy used by microsporidia to restructure the cytoskeleton of intestinal cells , involving relocalization of actin and reorganization of a structure called the terminal web , which may be a barrier to exit . In addition , we found that despite large numbers of parasites exiting out of intestinal cells , this process does not cause cells to burst . Our findings indicate that microsporidia , which are completely dependent on their hosts for replication , have evolved a regulated and non-damaging mechanism of exit that shares similarities with strategies used by evolutionarily distant bacterial pathogens . This study provides new insights into the methods by which pathogens restructure live intestinal cells to facilitate their spread and propagation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology",
"microbiology",
"host-pathogen",
"interaction"
] | 2011 | Non-Lytic, Actin-Based Exit of Intracellular Parasites from C. elegans Intestinal Cells |
Transposable elements ( TEs ) have been active in the mammalian genome for millions of years and the silencing of these elements in the germline is important for the survival of the host . Mice carrying reporter transgenes can be used to model transcriptional silencing . A mutagenesis screen for modifiers of epigenetic gene silencing produced a line with a mutation in Trim33; the mutants displayed increased expression of the reporter transgene . ChIP-seq of Trim33 in testis revealed 9 , 109 peaks , mostly at promoters . This is the first report of ChIP-seq for Trim33 in any tissue . Comparison with ENCODE datasets showed that regions of high read density for Trim33 had high read density for histone marks associated with transcriptional activity and mapping to TE consensus sequences revealed Trim33 enrichment at RLTR10B , the LTR of one of the youngest retrotransposons in the mouse genome , MMERVK10C . We identified consensus sequences from the 266 regions at which Trim33 ChIP-seq peaks overlapped RLTR10B elements and found a match to the A-Myb DNA-binding site . We found that TRIM33 has E3 ubiquitin ligase activity for A-MYB and regulates its abundance . RNA-seq revealed that mice haploinsufficient for Trim33 had altered expression of a small group of genes in the testis and the gene with the most significant increase was found to be transcribed from an upstream RLTR10B . These studies provide the first evidence that A-Myb has a role in the actions of Trim33 and suggest a role for both A-Myb and Trim33 in the arms race between the transposon and the host . This the first report of any factor specifically regulating RLTR10B and adds to the current literature on the silencing of MMERVK10C retrotransposons . This is also the first report that A-Myb has a role in the transcription of any retrotransposon .
Approximately half of the genomes of humans and mice are made up of transposable elements ( TEs ) and numerous pathways , both genetic and epigenetic , have evolved to repress their expression . Each TE has had a period of transpositional activity during which it spreads through the genome and this is dependent on its transcriptional activity in the germline . It has been suggested that during periods of epigenetic reprogramming , such as occurs in the germline , when DNA methylation is low , other mechanisms are used to suppress the expression of recently transposed retrotransposons [1] . Because of the difficulty of mapping individual repeats back to the genome , the epigenetic and transcriptional state of retrotransposons has been difficult to study . Improved next generation sequencing chemistry , resulting in longer reads , is helping to overcome this problem and bioinformatics tools are being refined to deal with repetitive elements . Unbiased genetic screens for modifiers of epigenetic gene silencing have been carried out in a number of model organisms , including the mouse , and have provided a valuable tool in the identification of the proteins responsible for mediating transcriptional silencing of inserted reporter transgenes [2 , 3] . The silencing of transgenes is thought to mimic , in many ways , that of retrotransposons [4 , 5] . We have carried out a dominant mouse screen using a line containing a GFP transgene array that undergoes stochastic silencing in erythroid cells and the alleles identified are termed Modifiers of murine metastable epialleles Dominant ( MommeDs ) [4] . The underlying genes responsible for many but not all of the MommeD lines have been identified and reported; MommeD1 , 2 , 4 , 5 , 8–10 , 12–14 , 16–19 , 23 , 27 , 28 , 30–40 , 42 [3] , MommeD6 , 8 [6] and the results of the screen have recently been reviewed [7] . The genes encode known epigenetic modifiers , including DNA methyltransferases , chromatin remodellers , histone methyltransferases , histone deacetylases and some previously uncharacterised proteins , such as Smchd1 and Rlf . Here we report , for the first time , the mutation underlying MommeD44 and the consequences of this mutation on the transcriptome in the mouse . MommeD44 heterozygous mutants showed an increase in the expression of the reporter transgene compared to that of their wildtype littermates . The underlying N-ethyl-N-nitrosourea-induced mutation was found to have produced a null allele of the gene Trim33 , Tripartite motif containing 33 , also known as Tif1γ , which codes for a protein that contains a tripartite motif with ubiquitin ligase activity and two chromatin binding domains; a bromodomain and a PHD ( plant homodomain ) finger . The function of Trim33 remains poorly understood . In cell culture , Trim33 can act as transcriptional coregulator via the TGFβ pathway [8] and mice homozygous for a null allele die during embryonic development [9] . While few functional studies have been carried out in whole animals , it has been shown that Trim33 can act in combination with the close homologues , Trim24 and Trim28; Lox-Cre knockdown of Trim33 in the liver of mice null for Trim24 results in increased expression of the VL30 class of retrotransposons [10] and hepatocellular carcinoma [11] . Unlike Trim28 , which is known to recruit chromatin modifying protein complexes to the 5’ UTR of ERVs [12] , the mechanism by which Trim33 acts has been unclear . Trim33 has been shown to be highly expressed in the spermatogonia and primary spermatocytes of the testis [13] but its role in these cells is unknown . We have carried out ChIP-seq in adult testis and show widespread binding to active gene promoters , as well as to the members of the RLTR10B class of retrotransposon that are marked by H3K27ac . Using publically available datasets , we find considerable overlap between Trim33 and A-Myb binding sites . Transcriptome analysis , using heterozygous MommeD44 mice , demonstrates a role for Trim33 as a dominant transcriptional repressor of RLTR10B elements in the testis and a combination of bioinformatics and biochemical studies suggests that this repression involves the A-Myb transcription factor . This is the first report of ChIP-seq for Trim33 and the first report of a function for Trim33 in the testis .
The experimental pipeline for the screen has been described previously [3] . The mutagenesis was carried out in the FVB/NJ background and the MommeD44 founder was identified because it showed an increase in the proportion of red blood cells expressing the GFP reporter ( Fig 1A ) . As such , MommeD44 was classified as a suppressor of variegation . This readout was used to maintain the MommeD44 allele for five generations prior to performing genetic linkage and exome sequencing . Mapping was carried out following a F2 backcross to Line3C ( a C57BL/6J strain carrying the same GFP transgene array at the same location ) , using the general method described previously[6] . Using the Illumina GoldenGate SNP genotyping assay , we identified an interval on Chr3 between 75Mb and 155Mb ( S1 Fig ) . Fine mapping , using additional mice , reduced the interval to Chr3 between 99Mb and 109Mb ( Fig 1B ) . To identify the underlying mutation , whole exome deep sequencing of DNA was carried out and variants were called within the linked interval . An exonic mutation was identified in the Trim33 gene , which is located in the linked interval . The A → T mutation changes a lysine to a premature stop codon ( Fig 1C ) . We designate this allele Trim33MommeD44 and genotyping mice from the colony showed that the presence of the mutation correlated well with the altered GFP profile ( S2 Fig ) . RTqPCR was carried out to determine the effect of the mutation on the level of Trim33 mRNA ( Fig 1D ) . The study was carried out in testis because Trim33 is expressed in that tissue at much higher levels than in any other tissue in the adult [13] . Heterozygotes have approximately half the Trim33 mRNA level seen in wildtypes , suggesting that the mutant transcript undergoes nonsense mediated mRNA decay . Using an antibody that binds specifically to Trim33 ( S3A Fig ) , a similar decrease in the level of Trim33 protein was found in heterozygous mutant testis ( S3B and S3C Fig ) and no band can be seen in homozygous mutant embryos ( S3D Fig ) . We examined the viability of homozygous mutants . As expected , embryos homozygous for the mutation were grossly abnormal at E10 . 5 and were not recovered after this stage ( Fig 1E ) . The ratios of heterozygous to wildtype offspring suggest no loss of heterozygotes . At three weeks , heterozygotes showed no difference in body weight when compared to wildtypes ( Fig 1F ) . The heterozygous males were fertile and sections from adult testis showed no obvious abnormalities ( S4 Fig ) . To identify the binding sites of Trim33 , across the genome , the same anti-Trim33 antibody ( Bethyl Laboratories A301-060A ) was used to perform ChIP in adult testis followed by high throughput sequencing of Trim33-bound DNA . Approximately 40 million reads were generated for both Trim33 chromatin immunoprecipitated DNA and an Input control and approximately 30 million of these uniquely aligned to the genome in both cases ( S1 Table ) . Peaks were called in the aligned sequence data using the program MACS [14] . This method looks for significant enrichments in the ChIP-seq data file when compared to the Input data file . A total of 9109 peaks were identified across the genome ( S2 Table ) . Trim33 enrichment was validated using ChIP qPCR at all loci tested ( four of four ) ( S5 Fig ) . The majority of the 9109 peaks occurred within 2Kb from a RefSeq transcription start site ( TSS ) ( Fig 2A ) . Furthermore , when read density was calculated across the gene body unit for all RefSeq genes , Trim33 was found to be enriched over the transcription start site ( Fig 2B ) . Using the ENCODE testis dataset , we found that regions of high read density for Trim33 had high read density for H3K4me3 and H3K27ac ( Fig 2C ) . These histone modifications are generally associated with actively transcribed genes and this has been reported specifically in spermatogonial cells [15] . There were a significant proportion of RefSeq genes with both H3K27ac and H3K4me3 that did not have Trim33 ChIP-seq peaks , suggesting that not all genes that are transcriptionally active in the testis are bound by Trim33 . Using the GREAT gene ontology tool [16] , the 9109 Trim33 peaks were interrogated for association with genes of particular function . The peaks were significantly enriched adjacent to genes that are classified by the MGI expression database as “the TS28_primary spermatocyte” ( Fig 2D ) . This is equivalent to an adult animal’s primary spermatocytes . This was expected , as the majority of Trim33 peaks were located at promoters with active histone marks and likely to be expressed in testis . Primary spermatocytes represent the majority of the cells of the testis . Trim33 was identified in a screen for genes involved in silencing of a GFP transgene , a locus that had been inserted into the genome . This raised the possibility that it might have a role in regulating other elements recently integrated into the genome , including LINE-elements and LTR-containing retrotransposons . Using the Trim33 ChIP-seq data , we mapped reads to consensus sequences for each repeat annotation from the rodent Repbase repository , which groups elements by sequence , e . g . IAPEZI , IAPLTR1_Mm , IAPLTR2b etc . Of the approximately 1600 repeat consensus sequences , we considered only the 309 that had a RPKM ( reads per kilobase per million ) >10 in testis ( S3 Table ) . Only 20 classes showed a fold change ( over Input ) of greater than or equal to 1 . 25 ( Fig 3A and S3 Table ) . We found that Trim33 was enriched almost seven-fold at RLTR10B and to a lesser extent ( two-fold ) at RLTR10B2 . These are among the youngest retrotransposon elements in the genome [17] and belong to the ERVK family of repetitive elements . Using the RepeatMasker locations of all RLTR10B and RLTR10B2 elements , we found that 35% of the former and 12% of the latter overlapped with Trim33 ChIP-seq peaks ( Fig 3B ) . Some reads were likely to have aligned poorly at individual RLTR10B elements due to their repetitive nature . After filtering out reads with a Bowtie2 mapping quality score of below 20 , a large proportion of reads still remained at these sites suggesting that they were uniquely mapped . Importantly , many reads overlapped with non-repeat derived flanking sequences to which short deep sequencing reads are more readily mapped ( Fig 3C ) . Fortuitously , ChIP qPCR validations , discussed in the previous section , included a ChIP-seq peak overlapping a RLTR10B element located in an intron of the gene Fgf2 ( S5 Fig ) . Trim33 lacks a classic DNA binding domain , suggesting other mechanisms are required to provide specificity . Studies in other tissues/cell types have found that Trim33 uses either Smad4 or β-catenin to anchor to DNA [8] . To address how Trim33 might be directed to DNA in testis , we used the MEME program to identify any consensus sequences from the Trim33 binding peaks with a high significance , P value < 1 . 0e-20 . These peaks ( N = 2338 ) generally had high enrichment over Input ( 99% were at least 5 fold enriched ) and these were the most likely to represent direct interactions between Trim33 and the transcription factors to which it binds . The three most highly enriched motifs are found in RLTR10B elements , as expected . Using the JASPAR CORE 2014 database of transcription factor DNA-binding sites , a match was found between an RLTR10B MEME-generated consensus sequence and the Myb consensus DNA binding site ( Fig 3D ) . The Myb consensus sequence emerged from 502 individual sites . Several other DNA binding factors were identified with lesser significance , however , no match was found with Smad or β-catenin binding sites ( S6 Fig ) . As ChIP-seq has been carried out for A-Myb ( also known as Mybl1 ) in testis [18] , we performed unsupervised hierarchical clustering of ChIP-seq read densities for Trim33 , A-Myb and two Encode histone marks , H3K27ac and H3K27me3 ( representing active and repressive chromatin , respectively ) at RLTR10B elements ( Fig 3E ) . We found good overlap between those RLTR10B elements that bind Trim33 , those that bind A-Myb and those enriched for H3K27ac . The repressive chromatin mark H3K27me3 was not enriched at RLTR10B elements . This lack of H3K27me3 enrichment at RLTR10B elements has been reported in ES cells [19] . Given the similar profiles of Trim33 and A-Myb at RLTR10B elements , we compared sites of Trim33 enrichment and those of A-Myb enrichment across the entire genome and found that approximately half of all A-Myb peaks overlapped with Trim33 peaks and approximately one fifth of Trim33 peaks overlap with A-Myb peaks ( Fig 3F ) . These findings suggest that A-Myb has a role in the actions of Trim33 across the genome in testis . RLTR10B elements were clustered by similarity of the Trim33 and A-Myb read density and those with a high read density for both were found to be enriched for the JASPAR database Myb consensus DNA-binding sequence ( S7A Fig ) . RLTR10B elements that bind Trim33 and A-Myb account for 78% ( 390/502 ) of the MYB sites identified in our Trim33 ChIP-seq motif analysis . RLTR10B with no enrichment for Trim33 and A-Myb were not enriched for the Myb DNA-binding sequence ( S7A Fig ) . This supports a role for A-Myb in the function of Trim33 in transcriptional regulation of RLTR10B-containing retrotransposons in the testis . Furthermore , several Trim33 ChIP-seq datasets have recently been carried out in lymphoid cell lines that do not express A-Myb [20] . Reanalysis of these datasets revealed little affinity of Trim33 for RLTR10B elements in these cell types ( S7B Fig ) . The RLTR10B consensus sequence has four tandem sites matching the JASPAR CORE database Myb consensus binding site located at the 5’ region of the RLTR10B consensus sequence ( S8 Fig ) . Among the broader category of 447 ERV2 elements , 12 contain the consensus Myb binding site but in all cases these are interspersed across the element ( S8 Fig ) . Given that Trim33 binds to chromatin and functions as a transcriptional repressor via its ubiquitin ligase activity [21] , it seemed likely that A-Myb is a testis-specific target of Trim33 ubiquitination . Two ubiquitination sites have been identified on the A-Myb protein , the amino acids K79 and K140 [22] . Combinations of tagged A-MYB , ubiquitin and TRIM33 were overexpressed in HEK293 cells; A-MYB ubiquitination was substantially increased in the presence of TRIM33 ( Fig 4A ) . Knock down of TRIM33 using shRNA decreased the level of ubiquitinated A-MYB ( Fig 4B ) . To demonstrate the functional significance of this , A-MYB levels were monitored in the presence of cycloheximide , which prevents protein translation . In the presence of ubiquitin alone , A-MYB survived to 10 hours whereas in the presence of ubiquitin and TRIM33 , the levels of A-MYB survived to only 6 hour time point ( Fig 4C ) . In an attempt to identify an interaction between the co-expressed TRIM33 and A-MYB , we carried out immunoprecipitation for A-MYB ( using the FLAG tag ) followed by Western blotting for TRIM33 ( using the GFP tag ) . TRIM33 could be detected following immunoprecipitation with anti-FLAG ( A-MYB ) but not following immunoprecipitation with anti-IgG ( Fig 4D ) . In the absence of a spermatogonial cell line , immortalized MEFs were used to measure the expression of RLTR10B elements following overexpression of A-MYB and ubiquitin . Overexpression of A-MYB resulted in a small increase in RLTR10B expression and this effect was reversed after co-expression with ubiquitin ( S9 Fig ) , consistent with our hypothesis . We were keen to identify loci that might be sensitive to haploinsufficiency of Trim33 in the testis and that overlapped with Trim33 enrichment . We purified RNA from the testis of wildtype ( n = 3 ) and heterozygous Trim33MommeD44/+ ( n = 4 ) mice and carried out RNA-seq . Using the R package DESeq and Ensembl transcript annotations ( NCBI build 37 ) , we identified 39 genes that were significantly differentially expressed ( Fig 5A and S4 Table ) . Trim33 was ranked third in this list and was 0 . 68 fold down-regulated , close to expectation . The majority of the changes showed increased expression levels in the mutants , consistent with a role for Trim33 as a repressor of transcription in this tissue . Using RTqPCR we were able to validate the majority of those tested ( 9 of 11 ) ( S10 Fig ) . However , only eight of the 39 showed overlap with Trim33 binding peaks . Interestingly , the most significantly deregulated gene , Nmnat3 , upregulated in the mutants by two-fold , and validated by RTqPCR ( S10 Fig ) has an upstream RLTR10B element bound by Trim33 ( Fig 5B ) . In the testis , reads for the Nmnat3 gene start at this element and read across into exon2 ( the first coding exon ) showing that transcription initiates in the repeat element ( S11 Fig ) . Clonal bisulphite sequencing inside this RLTR10B showed no change in DNA methylation in testis of MommeD44 heterozygotes compared to that of wildtype mice ( S12 Fig ) , despite a two-fold change in transcription . We used Cufflinks to establish transcript annotations guided by both the raw RNA-seq data and existing annotations for the mouse mm9 genome to identify other upstream RLTR10Bs that initiate transcription of an adjacent gene . This set was used to identify transcripts that were significantly differentially expressed and linked to a RLTR10B or RLTR10B2 . Ten were found ( S5 Table ) , including Nmnat3 . A comparison was made between the Trim33 ChIP-seq data mapped against the consensus sequences for each repeat class , described above , and the RNA-seq mapped against the same consensus sequences ( Fig 5C ) . The RLTR10B and RLTR10B2 elements both deviate from all other elements with respect to Trim33 binding , as expected . A slight increase can be seen in expression of the consensus RLTR10B element in heterozygotes compared to wildtypes ( fold change = 1 . 33 , p value adjusted for multiple testing = 0 . 079 ) ( Fig 5C ) . We did not expect a large increase in expression of this class in the RNA-seq data as the majority of RLTR10B elements do not have A-Myb sites and are not bound by Trim33 ( S7A Fig ) . This increase was validated using RTqPCR with primers designed over the Myb consensus binding sequences but not with primers designed elsewhere in the RLTR10B consensus sequence ( S13A Fig ) , suggesting that the changes seen are limited to those RLTR10B elements with multiple Myb binding sites at the 5’ end . A second individual transcript which was upregulated in heterozygous MommeD44 mice and overlapping with an RLTR10B element , the Vmn1r181 gene ( S5 Table ) , was validated using two primers; one annealing upstream of the normal reading frame in the ectopic transcript and the other inside the normal gene reading frame ( S13B Fig ) . RLTR10B expression was measured in E9 . 5 embryos , in which A-Myb is not expressed . No change was seen across wildtype , heterozygous or homozygous MommeD44 embryos , consistent with an A-Myb independent function for Trim33 at RLTR10B in whole embryos ( S13C Fig ) . Interestingly , Nmnat3 has been shown to be down regulated ten-fold in the testis of mice with reduced A-Myb function [18] . Using their raw RNA-seq data and the consensus RLTR10B sequence , expression of RLTR10Bs was found to be decreased two-fold in mice heterozygous for a mutation in A-Myb and ten-fold in mice homozygous for the mutation ( S14 Fig ) . These findings are consistent with a central role for A-Myb in transcriptional activity of the RLTR10B elements .
Transposable elements have been a driving force in the structure and evolution of the mammalian genome . Deep sequencing of 17 mouse genomes has revealed over 100 , 000 transposable element variants that have survived selection over the past 2 million years of Mus lineage evolution [23] . From an inferred evolutionary history of TE family activity , ERVs and in particular ERVKs , appear to be expanding rapidly in the mouse [24] . The ERVK class , also known as ERV2 , contains IAPs , Etn/MusD and RLTR10B-containing retrotransposons such as MMERVK10C . The vast majority of the expansion of LTR retrotransposons is thought to have occurred in the paternal germline [23] . This is consistent with findings from a study in which IAP-GFP transgenes were inserted into the mouse genome . Expression of these transgenes was found to be limited to the male germline [25] . Most studies that have investigated the transcriptional regulation of the ERV2 class of repeats have focussed on piRNA pathways [23] . Directed mutagenesis in mice has identified nine proteins with a key role in protecting the male germline against retrotransposition of IAPs , Etn/MusD and MMERVK10C elements; Dnmt3L , Dnmt3a , Miwi2 , Mili , Mael , Tdrd1 , Tdrd9 , Gasz , Tex19 . 1 and most function in the piRNA pathway [23] . Several groups have reported on factors that silence MMERVK10Cs in the germline [26 , 27] but the mechanisms of silencing of RLTR10Bs has not been understood prior to our studies . Here we describe the first factor involved in silencing of the RLTR10Bs elements . Trim33 has been identified in a number of screens designed to find genes involved in transcriptional silencing . It was found in a RNA inhibition ( RNAi ) screen carried out in a human cell line to find proteins required for transgene silencing [28] . The fly homologue of Trim33 , Bonus , was identified as a suppressor of variegation ( i . e . reduced levels of Bonus result in an increased proportion of cells expressing the reporter transgene ) when tested for its ability to alter transcription of a variegating reporter locus [29] . Here we report , the identification of Trim33 in a mutagenesis screen for modifiers of transcriptional silencing in the mouse . In this screen it also behaves as a suppressor of variegation . Trim33 was identified in a zebrafish screen for genes involved in the development of the haematopoietic system [30 , 31] . It is known to be required for development of ectoderm in Xenopus [32] . Knockout of Trim33 in the mouse has been shown to result in embryonic lethality as a consequence of excessive TGFβ/Nodal signalling [33] . Our findings of embryonic lethality of embryos homozygous for the TrimMommed44 allele are consistent with this report . Trim33 is not thought to bind DNA directly and available evidence suggests that its transcriptional effects occur via DNA-binding cofactors . For example , the role of Trim33 in TGF-beta signalling in cell lines has been shown to involve the Smad transcription factors that bind DNA in a sequence-specific manner [8 , 34] . [It has been suggested that the mechanism by which Trim33 inhibits TGF-beta signalling during the development of ectoderm in Xenopus involves the ubiquitination of Smad4 [32] . The model proposes that Trim33 inhibits expression of the target locus by destabilising the Smad cofactor . ] The role of the chromatin binding domains ( PHD and Bromodomain ) of Trim33 remains unclear . Previous studies , following overexpression of the chromatin binding domains in somatic cell lines , mapped the binding specificity to H3K4me0 and acetylated lysine residues [8 , 21] . However , our ChIP-seq studies , carried out in vivo , found that most Trim33 peaks in testis overlap with peaks for H3K4me3 and H3K27ac and presumably not for H3K4me0 . Others have suggested that the PHD and Bromodomain of Trim33 are required to activate its ubiquitin ligase activity [21] . Multiple mouse models of cancer have demonstrated a role for Trim33 as a tumour suppressor [11 , 35 , 36] . Trim33 was found to act as a tumour suppressor in the pancreas of mice and humans [35] . The mechanism underlying the tumour suppressor function is thought to be Smad4-independent [37] . Recent studies have found that TRIM33 abolishes tumour cell proliferation and tumorigenesis by degrading nuclear β-catenin via ubiquitination [36] . Taken together , these findings indicate Trim33 can be considered a corepressor of transcription that functions by ubiquitinating DNA-binding cofactors and that it has important roles in development and cancer . Trim33 is expressed an order of magnitude higher in the testis than in any other adult mouse tissue [13] . It is expressed in spermatogonia , preleptotene spermatocytes and round spermatids but its function in these cells is unknown . It is also expressed in one class of the somatic cells in the testis , the Sertoli cells , but these are much less abundant than the germ cells . Since it has been identified in a number of screens for gene silencing and since the silencing of retrotransposons is a critical function of the germline , we hypothesised that it might have a role in this process . Our study is the first report of ChIP-seq for Trim33 in any tissue . We show that Trim33 binds many active gene promoters in the testis ( i . e . those promoters marked by H3K4me3 and H3K27ac ) . Trim33 also shows enrichment at RLTR10B elements , a subgroup of ERVK LTR retrotransposons , consistent with our hypothesis . RLTR10Bs are among the chronologically youngest retrotransposable elements in the mouse genome , ranked 5th out of 546 elements analysed [17] . The only groups considered younger than RLTR10Bs are IAPs and these are known to have retrotranspositional activity in the mouse [38] . In general , the youngest retrotransposable elements will have accumulated less defects ( i . e . mutation of DNA binding sites ) and will have retained the ability to be expressed . Knowing that haploinsufficiency for Trim33 was sufficiently disruptive to increase expression of the reporter transgene in our mutagenesis screen , we were keen to test whether it would alter the expression of other genes in the testis . RNA-seq suggests no dramatic effect across the genome but confirms a role for Trim33 in silencing RLTR10B . The fact that the transgene and the RLTR10B elements , but not other Trim33-bound loci , are sensitive to Trim33 dosage is not understood . It might be that other Trim family members ( and other factors ) can compensate for reduced levels of Trim33 at most loci . It is likely that in the complete absence of Trim33 , many of the Trim33 bound loci would be affected but the early demise of the homozygous embryo precludes testing . We have identified a role for Trim33 in binding and silencing the RLTR10B-containing class of retrotransposon in testis . Many other groups of retrotransposons , including active IAP elements , are enriched for the same histone marks found on RLTR10B elements [19] but are not bound by Trim33 , suggesting that Trim33 does not bind via its bromodomain or PHD finger . On the assumption that Trim33 binds DNA via a transcription factor , we searched for a transcription factor consensus binding site and found four separate Myb binding sites at the 5’ end of the RLTR10B consensus sequence and the individual RLTR10B upstream of Nmnat3 . This tandem arrangement has previously been shown to increase the affinity for Myb binding to DNA [39] A-Myb is a master regulator of male meiosis and is expressed specifically during spermatogenesis [40] but has not previously been reported to bind retrotransposons . Given the overlapping ChIP-seq signal between A-Myb and Trim33 at these repeats ( Fig 3E ) , it is likely that Trim33 functions to repress transcription at these sites via A-Myb . The Momme screen reporter transgene contains an AACTGTCT element in the HS40 alpha globin hypersensitive site and this fits the MYB binding consensus site–AACTG ( C/T ) C ( A/T ) . A-MYB has been shown to act as a transcriptional activator in peripheral blood cells [41] . It is reasonable to suggest that Trim33 acts as a transcriptional repressor of the reporter transgene via ubiquitination of A-Myb ( or another target of Trim33 ubiquitination ) , although we have no direct evidence of this . Myb family members are subject to several types of post-translational modifications , including phosphorylation , acetylation , and ubiquitination . Furthermore , the ubiquitination of B-Myb and C-Myb has been shown to inhibit the transcriptional activation functions of these two factors [42 , 43] . Trim28 , like Trim33 , can ubiquitinate proteins and has been shown to bind to C-Myb and repress its ability to function as a transcriptional activator [44] . Here , we have demonstrated that Trim33 can ubiquitinate A-Myb and that this regulates the abundance of A-Myb , probably by the ubiquitination–proteasome pathway as has been suggested for Trim33 mediated degradation of β-catenin [36] . Publically available datasets show that reduction in A-Myb in mouse testis results in a decrease in expression of Nmnat3 [18] . We have reanalysed their datasets to search for effects on repeats and find dramatically decreased expression of RLTR10Bs ( S14 Fig ) . In addition , the cell types in the adult testis that specifically express high levels of Trim33 overlap with those that express high levels of A-Myb [40] . Given that Trim33 can ubiquitinate and regulate the abundance of A-Myb , a simple model of the mechanism by which Trim33 silences RLTR10B elements in the germline would include the ubiquitination of A-Myb ( Fig 6 ) . Transcriptional activity at retrotransposons is species specific , consistent with rapid evolution of retroviral subtypes [23] . It is likely that a subset of RLTR10Bs have recently evolved Myb DNA binding sites to capitalise on the critical role that the A-Myb transcription factor has in gene expression in germ cells in order to ensure their continued retrotransposition . Suppression of A-Myb by Trim33 provides a plausible mechanism by which the host keeps retrotransposition in check .
The ENU screen was carried out in Line3 , a transgenic line that is homozygous for the GFP transgene and on an FVB/NJ inbred background , as previously described [4] . Maintenance of the MommeD44 allele was carried out on the Line3 background . The Line3C was used for linkage studies and was produced by crossing Line3 to C57BL/6J for 10 generations and selecting for homozygosity of the GFP transgene . All breeding crosses and experimental procedures using the MommeD44 allele were carried out using mice at least 5 generations from the founder mouse . Procedures were approved by the Animal Ethics Committee of LaTrobe University , under approval numbers AEC 12–74 and AEC 12–75 . All mice sacrificed in the study were anesthetized using isoflurane and euthanised by cervical dislocation . The expression vector encoding A-MYB-FLAG was purchased from OriGene ( Rockville , MD , USA ) . TRIM33-GFP was obtained from DR Kyle Miller ( Addgene plasmid #65399 ) . Knockdown shRNA for Trim33 were obtained from Dr Joan Massague ( Addgene plasmid # 15728 ) . PCR was carried out using DNA from tail tips using the following conditions: 94 degrees Celsius 6 minutes , 30x cycles with 94 degrees Celsius 30 seconds , 60 degrees Celsius 45 seconds 72 degrees Celsius 45 seconds , followed by a final annealing step of 72 degrees Celsius 6 minutes . All primers used in the study are described in S6 Table . Flow cytometry of blood from 3 week old mice collected in FACSFlow Sheath Fluid ( BD Biosciences ) was carried out and analysed on a Guava easyCyte HT ( Merck/Millipore , Darmstadt , Germany ) and with the Guava InCyte software , respectively . Erythrocyte green fluorescence ( 525nm ) was recorded and a GFP-positive gate was set to exclude 99% of wildtype erythrocytes . MommeD44 heterozygous mice were backcrossed twice to Line3C ( see above ) and phenotyped for GFP expression by flow cytometry . DNA from tail tissue collected during flow cytometry procedures was used to perform linkage analysis . The Illumina GoldenGate genotyping assay ( Mouse Medium Density Linkage Panel ) was used with 10 wildtype and 13 heterozygous mice . MommeD44 wildtype samples should only have heterozygous C57BL/6J SNPs surrounding the causative mutation and MommeD44 mutants should have FVB and C57BL/6J SNPs at this interval . The Mouse Medium Density Linkage panel contains 766 measurable SNPs between C57BL/6J and FVB/NJ . Samples were genotyped following the Illumina protocol and genotype calls were made using the Genotyping module of the GenomeStudio v1 . 1 software . Only samples with a call rate >95 were accepted . The linked interval was identified based on a peak in the LOD score . Fine mapping was carried out using primers amplifying C57BL/6J or FVB SNP loci that could be cut with restriction enzymes to determine genotype . Of the 95 F2 backcrosses , 8 mice had SNP profiles that were inconsistent with the mapping and were excluded . This small ( <10% ) error rate in phenotyping is commonly encountered in this screen [4] . Exome capture was performed using the Roche NimbleGen reagents ( SeqCap EZ Mouse Exome , version Beta 2 , 110603_MM9_exome_rebal_2EZ_HX1 , Madison , WI , USA ) as per the Illumina optimized Roche NimbleGen SeqCap User’s Guide ( version 1 . 0 ) and using a Bioruptor ( Diagenode , Liège , Belgium ) for DNA fragmentation . Libraries were sequenced using the illumina GAIIx platform and reads were aligned to the mouse ( build 37 , mm9 ) genome using bwa aln and bwa sampe programs [45] . For further details see [3] . As control sequence , that lacked MommeD44 ENU mutations , other MommeD lines previously identified in the screen and described elsewhere were used [3] . Varscan output was scanned manually for likely heterozygous mutations that could be validated in additional MommeD44 mutant mice . Sanger Sequencing or restriction enzyme digest with the enzyme Mse1 ( New England Biolabs ) that specifically cuts at the ENU mutation site in the Trim33MommeD44 allele was used . Timed matings between heterozygous mutant females and heterozygous mutant males were set up and the detection of a vaginal plug was counted as 0 . 5 dpc . Genotyping was carried out using DNA extracted from embryos . Bisulphite conversion was carried out on 1ug of DNA extracted from whole testis using the EpiTect Bisulphite Kit ( Qiagen , Doncaster , VIC , Australia ) according to the manufacturer’s instructions . The bisulphite conversion rate was at least 99% and sequences were analysed using the BiQ Analyser software . Oligonucleotides are provided in S6 Table . PCR cycling conditions were as follows: 95 degrees Celsius 10 minutes , 30x cycles with 95 degrees Celsius 15 seconds 55 degrees Celsius 15 seconds , 72 degrees Celsius 30 seconds . PCR products were cloned using a pGEM-T Easy Vector ( Promega , Alexandria , NSW , Australia ) and sequenced using The BigDye Terminator v3 . 1 Cycle Sequencing Kit ( Life technologies , Mulgrave , Victoria , Australia ) as per kit instructions . HEK293T and mouse embryonic fibroblasts were cultured in DMEM supplemented with 10% fetal calf serum ( Invitrogen , Scoresby , VIC , Australia ) at 37ᴼC in a humidified 10% CO2 incubator . For transient expression of cDNA or RNAi vectors , HEK293 or MEF cells were transfected with expression plasmids using the Fugene 6 reagent ( 11814443001 , Roche , Castle Hill , NSW , Australia ) following the manufacturer’s instructions . For A-MYB survival assays , cells were transfected with indicated plasmids for 48 hours and then treated with cycloheximide to a final concentration of 10 ug/ml . Cells were lysed at indicated time points and A-MYB protein levels were assessed by Western blotting . Whole-cell extract was prepared from testis of adult ( 12 week old ) mice in 8M urea lysis reducing buffer ( 8 M urea , 1/10 vol . glycerol , 1/20 vol . 20% SDS , 1/2 , 000 vol . 1 M dithiothreitol , 1/100 vol . 1 M Tris , pH 6 . 8 ) . For ubiquitination assays , HEK293T cells were lysed in RIPA buffer ( Tris 50mM , NaCl 150mM , NP-40 1% , DOC 0 . 5% , SDS 0 . 1% ) , diluted 1:10 in ONYX buffer ( Tris 20mM , NaCL 135mM , MgCl2 1 . 5mM , EGTA 1mM , Triton X-100 1% , Glycerol 10% ) and incubated overnight with anti-FLAG antibody at 4ᴼC . Following incubation , antibodies were washed in ONYX buffer and proteins were eluted in 8M urea lysis buffer . For protein co-immunoprecipitations the Nuclear Complex Co-IP Kit ( 54001 , Active Motif , Carlsbad , CA , USA ) was used as per manufacturer’s instructions . BCA ( Thermo Scientific , Waltham , MA , USA ) was used to quantify protein and total lysates were separated according to size on polyacrylamide gels ( Biorad , Gladesville , New South Wales , Australia ) . Antibodies used for western blotting were rabbit polyclonal anti-Trim33 ( A301-060A , Bethyl Laboratories , Montgomery , TX , USA ) , rabbit polyclonal anti-A-Myb ( HPA008791 , Sigma Aldrich , Castle Hill , NSW , Australia ) , mouse monoclonal anti-GFP ( 11814460001 , Roche , Castle Hill , NSW , Australia ) , mouse monoclonal anti-HA ( 6E2 , Cell Signaling , Boston , MA , USA ) , mouse monoclonal anti-Gapdh ( MAB374 , Merck/Millipore , Darmstadt , Germany ) mouse monoclonal anti-HSP70 ( MA3-028 , Scoresby VIC , Australia ) and rabbit polyclonal anti-γ-Tubulin ( T5192 , Sigma Aldrich , Castle Hill , NSW , Australia ) . Anti-Flag was a gift from Dr Lorraine O'Reilly at the Walter and Eliza Hall Institute of Medical Research . Antibodies used for immunoprecipitation were bead-conjugated anti-FLAG ( M8823 , Sigma Aldrich , Castle Hill , NSW , Australia ) and anti-IgG ( sc-2345 , Santa Cruz Biotechnology , Dallas , TX , USA ) . Total RNA was extracted from snap frozen tissues or cells using TRIzole reagent ( Life technologies , Mulgrave , Victoria , Australia ) according to manufacturer instructions . cDNA synthesis was carried out from total RNA using the QuantiTect Reverse Transcription Kit ( Qiagen , Doncaster , VIC , Australia ) and RTqPCR was performed with the QuantiTect SYBR Green reagent ( Qiagen , Doncaster , VIC , Australia ) with primers designed to span exon junctions in mRNA ( S6 Table ) . Samples were run on the CFX384 Touch Real-Time PCR Detection System ( Biorad , Gladesville , New South Wales , Australia ) , with the following conditions: 95 degrees Celsius 10 minutes , 39x cycles with 95 degrees Celsius 15 seconds then 60 degrees Celsius 1 minute , with a final step of 95 degrees Celsius 15 seconds . Each experimental sample consisted of three technical replicates and reverse transcriptase negative samples; a melt curve analysis was carried out after each run to confirm unique PCR product amplification . Relative cDNA abundance was calculated using the delta delta CT method normalizing to housekeeper gene expression indicated in the figures . Statistical analysis was performed using Student’s t test . RNA sequencing was carried out from total RNA submitted to the Australia Genome Research Facility ( AGRF , Parkville , Victoria , Australia ) . At least 20 million 100bp single end reads were generated on an Illumina HiSeq platform for each sample , using libraries generated using the illumina TruSeq RNA Sample Preparation kit ( Illumina , San Diego , CA , USA ) . Initial QC was performed by AGRF . Reads were aligned to the mouse genome ( NCBI 37 , mm9 ) using the program Tophat ( version 2 . 0 . 11 ) [46] with the following parameters: -I 100000—library-type = fr-unstranded—read-edit-dist 3—no-coverage-search—read-mismatches 3 . Read counts for gene exons were extracted using the program htseq-count ( version 0 . 6 . 1 ) [47] with the options -s no -m intersection-strict and using gene annotations from Ensembl ( release 67 ) . Differential gene expression was assessed using the R-package DEseq [48] , with default parameters . Genes were considered differentially expressed when an adjusted p value of at least 0 . 05 . Where indicated in the text , the program Cufflinks ( version 2 . 2 . 1 ) [49] was used to estimate differential gene expression of transcripts by creating annotations based on mapped reads . Mapped RNA-seq reads ( above ) were used to create transcript annotations using default settings for each sample , these were merged and then used to estimated differential expression of transcripts . Testis tissue from three 12 week old wildtype mice were snap-frozen and sent to Active Motif for ChIP , library preparation , sequencing and initial data analysis . The rabbit polyclonal anti-Trim33 ( A301-060A , Bethyl Laboratories , Montgomery , TX , USA ) was used for ChIP . Sequencing was carried out for 75mer read lengths on the NextSeq 500 platform ( Illumina , San Diego , CA , USA ) . Reads were aligned using the BWA program [45] with default settings . Peak calling was done with MACS ( Version 1 . 4 . 2 ) [14] by first filtering out duplicate reads and reads with a mapping quality of less than 25 , then using default parameters and the following options -s 75—bw 200 -m 10 30 –p 0 . 0000001 ( S1 Table ) . All heat plot and read tag density figures were generated using the seqMiner program ( version 1 . 3 . 3 ) [50] using default parameters and ChIP-seq data aligned to the mouse genome ( NCBI 37 , mm9 ) with the program Bowtie2 ( version 2 . 2 . 2 ) [51] with default settings . Heat plots were generated by subsampling all datasets to approximately 16 million reads . In the case of publically available ENCODE datasets , aligned reads in bam format were downloaded and were subsampled to approximately 16 million reads or in the case of H3K27ac , two biological replicates were combined to generate a data file with 16 million reads . The data sets supporting the results of this article are available in the Gene Expression Omnibus ( GEO ) repository , [GSE68617] . Motif discovery and enrichment was performed with highly significant Trim33 peaks ( P value <1 . 0d-20 , region summit +- 650bp ) using the MEME-ChIP ( version 4 . 10 . 0 ) [52] and MEME suite programs MEME , DREAME , CentriMo and Tomtom with default settings . Gene ontology analysis was carried out using Trim33 peak locations across the genome with the GREAT tool ( version 2 . 0 ) [16] with default settings . To estimate enrichment of repeat elements for ChIP-seq and RNA-seq datasets reads were mapped to a repeat assembly file containing a single FASTA entry for each repeat type defined in the rodent repeat sequences RepBase database [53] ( update 20 . 02 ) . The Bowtie2 ( version 2 . 2 . 2 ) [51] aligner was used to map reads aligning to each FASTA entry using default settings and RPKM values were extracted from the number of reads aligned at each entry and the library size for each data file . The data sets supporting the results of this article are available in the NCBI Gene Expression Omnibus under the accession code GSE68617 . | Almost half of the genomes of humans and mice are made up of transposable elements . During host evolution , subsets of these elements have periods of transpositional activity during which they spread throughout the genome . This is dependent on the transcriptional activity of these elements in the cells that contribute to the germline . Hosts have evolved pathways to silence their expression . A number of Trim family proteins have been found to have a role in silencing transposable elements , and it was previously shown that Trim33 shared this function in liver . However , the function of Trim33 in other tissues is poorly understood . Here we report a role for Trim33 in silencing a specific subset of retrotransposons that contain RLTR10B LTRs , in the germline . We also show the transcription factor , A-Myb , is responsible for activating transcription of these elements and it is likely that a subset of RLTR10Bs have recently evolved Myb DNA binding sites to capitalise on the critical role that the A-Myb transcription factor has in germ cells . Suppression of A-Myb activity by Trim33 provides a plausible mechanism by which the host keeps transposons in check . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Trim33 Binds and Silences a Class of Young Endogenous Retroviruses in the Mouse Testis; a Novel Component of the Arms Race between Retrotransposons and the Host Genome |
Quorum sensing ( QS ) is a mechanism of chemical communication that bacteria use to monitor cell-population density and coordinate group behaviors . QS relies on the production , detection , and group-wide response to extracellular signal molecules called autoinducers . Vibrio cholerae employs parallel QS circuits that converge into a shared signaling pathway . At high cell density , the CqsS and LuxPQ QS receptors detect the intra-genus and inter-species autoinducers CAI-1 and AI-2 , respectively , to repress virulence factor production and biofilm formation . We show that positive feedback , mediated by the QS pathway , increases CqsS but not LuxQ levels during the transition into QS-mode , which amplifies the CAI-1 input into the pathway relative to the AI-2 input . Asymmetric feedback on CqsS enables responses exclusively to the CAI-1 autoinducer . Because CqsS exhibits the dominant QS signaling role in V . cholerae , agonism of CqsS with synthetic compounds could be used to control pathogenicity and host dispersal . We identify nine compounds that share no structural similarity to CAI-1 , yet potently agonize CqsS via inhibition of CqsS autokinase activity .
Quorum sensing ( QS ) is a cell-cell communication process that enables bacteria to monitor population density and control behaviors as collectives . QS relies on the production , release , accumulation , and detection of extracellular signal molecules called autoinducers . At low cell density ( LCD ) , when autoinducer concentration is low , bacteria carry out behaviors that are successful when they act as individuals . At high cell density ( HCD ) , in response to autoinducer accumulation and detection , bacteria launch programs of gene expression that underlie group behaviors . Often , QS systems integrate information encoded in multiple autoinducers , which presumably enables bacteria to monitor numbers of kin ( intra-species QS ) , related family members ( intra-genera QS ) , and non-kin ( inter-species QS ) in the vicinal community [1 , 2] . The pathogen and model QS bacterium Vibrio cholerae has two known QS autoinducers: cholera autoinducer-1 ( CAI-1 , ( S ) -3-hydroxytridecan-4-one ) [3] and autoinducer-2 ( AI-2 , ( 2S , 4S ) -2-methyl-2 , 3 , 3 , 4-tetrahydroxytetrahydrofuran borate ) [4] ( Fig 1 ) . CAI-1 and AI-2 are produced by the CqsA [5] and LuxS [6 , 7] synthases and detected by the trans-membrane bound histidine sensor kinase receptors CqsS and LuxQ , respectively . LuxQ functions in conjunction with the periplasmic binding protein LuxP [8 , 9] . Two other QS receptors , VpsS and CqsR , have recently been identified but their ligands remain unknown [10 , 11] . All four QS receptors are hybrid two-component sensor histidine kinases [12–14] . The V . cholerae system functions as follows ( see Fig 1 ) : at LCD , in the absence of autoinducers , CqsS and LuxPQ act as kinases and shuttle phosphate , via the phospho-relay protein LuxU , to the transcription factor called LuxO [15–17] . LuxO~P activates expression of genes encoding a set of small regulatory RNAs called Qrr1-4 [18] . Qrr1-4 post-transcriptionally control the fates of the mRNAs encoding the two QS master transcription factors AphA and HapR . Specifically , the Qrr sRNAs promote production of the LCD master regulator , AphA , and they repress production of the HCD master regulator , HapR [19] . Thus , under this condition , bacteria enact behaviors appropriate for the individual lifestyle . At HCD , CAI-1 and AI-2 bind to and agonize CqsS and LuxPQ , respectively , converting them from kinases to phosphatases . LuxO is dephosphorylated , which inactivates it [20 , 21] . The Qrr sRNAs are not made , and thus , AphA production is not activated and HapR production is not repressed . HapR controls the expression of genes necessary for group behaviors . Of note , HapR represses expression of genes required for virulence factor production and biofilm formation [9 , 22–24] . Therefore , strategies that induce the V . cholerae HCD state render the pathogen avirulent [9 , 11 , 25 , 26] . Indeed , successful synthetic manipulation of V . cholerae QS through inhibition of LuxO has already been described [27] . Furthermore , commensal E . coli engineered to produce CAI-1 in the mouse intestine dramatically reduce cholera toxin production , and consequently , the lethality of V . cholerae infection [28] . LuxS and AI-2 are broadly distributed among bacteria and AI-2 is proposed to function in inter-species communication [4 , 29] . By contrast , CAI-1 and CqsS are almost exclusively restricted to vibrios suggesting they are used for intra-genus communication [3 , 30] . In V . cholerae , the CAI-1-CqsS system plays the dominant role relative to the weaker AI-2-LuxPQ system in directing V . cholerae QS-regulated gene expression , at least under laboratory conditions [9] . Here , we explore natural and synthetic modulation of the V . cholerae CqsS receptor to probe its overarching role in QS and to explore possibilities for synthetic manipulation of the process . We find that there exist ~40 CqsS dimers per cell at LCD , and QS induces CqsS production , which increases CqsS levels to ~170 dimers per cell during the LCD to HCD transition . By contrast , LuxQ levels remain unchanged as cells transition into QS-mode . The consequence of increased production of CqsS during the QS transition in V . cholerae is to enhance CAI-1 detection , which , as a result , dampens the response to AI-2 . Indeed , phosphatase-active LuxQ receptors are incapable of overriding kinase-active CqsS receptors . Conversely , phosphatase-active CqsS receptors can supersede kinase-active LuxQ receptors . We show that this feature of the network has important implications for the lifecycle of V . cholerae . Specifically , the hapA gene , encoding the dissemination-promoting protease HapA , is highly upregulated in response to CAI-1 but not in response to AI-2 . Thus , V . cholerae is capable of CAI-1-specific responses even though CqsS and LuxPQ signal through a common pathway . With respect to CqsS function , we identify two classes of synthetic CqsS agonists . Remarkably , although CqsS exhibits exquisite selectivity for CAI-1 when presented with CAI-1 derivatives [31] , responding to only a restricted subset of closely related CAI-1 analogs [30] , the synthetic compounds we identify , while capable of potently agonizing CqsS , share no structural similarity to CAI-1 . The synthetic agonists inhibit CqsS kinase activity , analogous to the mechanism by which CAI-1 functions to control CqsS signal transduction . The synthetic compounds display superior potency and efficacy compared to CAI-1 making them promising leads for V . cholerae prophylactics and/or therapeutics .
V . cholerae integrates sensory information from four receptors into a common , downstream QS pathway [9 , 11] . This network architecture makes it difficult to understand how individual autoinducers could produce unique QS responses . Here , we explore whether feedback operates in V . cholerae QS which , if so , could provide a possible route to autoinducer-specific responses . We first examined if either or both of the established V . cholerae receptors is controlled by QS by measuring CqsS and LuxQ levels and activity over growth . We describe in detail the strategy for assessment of CqsS . We used an exactly analogous strategy for analysis of LuxQ . We fused a 3XFLAG epitope to the C-terminus of CqsS and used this construct to replace the native cqsS gene on the V . cholerae chromosome . We verified that the CqsS::3XFLAG fusion functioned properly by monitoring its ability to control the density-dependent expression of luciferase , a heterologous QS reporter ( Fig 2A ) . The V . harveyi luxCDABE operon is frequently used as a convenient readout of QS-controlled gene expression in V . cholerae . Indeed , similar to WT V . cholerae , in the V . cholerae strain carrying CqsS::3XFLAG , maximal light production occurred following overnight growth ( i . e . , at HCD ) . Light output declined precipitously following dilution of the V . cholerae cells into fresh medium . Dilution reduces the autoinducer concentration to below the threshold required for detection , thereby transitioning the cells into LCD mode . During subsequent growth , autoinducers once again accumulate and are bound by their cognate receptors . Signal transduction activates lux expression , and light production commences . The canonical “U” shaped light production curve is the hallmark QS behavior ( Fig 2A ) . Thus , CqsS::3XFLAG functions like WT CqsS . We also deleted the genes encoding the other three QS receptors ( ΔcqsR ΔvpsS ΔluxQ ) and verified that , like WT CqsS , CqsS::3XFLAG could control gene expression when it was the only QS receptor present ( S1A Fig ) . We exploited the strain containing the CqsS::3XFLAG construct on the chromosome to examine whether CqsS levels change during QS transitions . Quantitative western blotting ( Fig 2B ) and the corresponding analysis of the data ( Fig 2C ) show that CqsS::3XFLAG increases from ~40 dimers per cell at LCD ( OD600 = 0 . 2 ) to ~170 dimers per cell at HCD ( OD600 = 2 . 0 ) . We could detect an increase in CqsS::3XFLAG at OD600 = 0 . 7 , immediately prior to induction of light production ( Fig 2A and 2C ) and CqsS levels steadily increased throughout growth . By contrast , similar measurements examining a functional LuxQ::3XFLAG fusion ( S1B Fig ) show that LuxQ::3XFLAG levels remain relatively unchanged throughout growth ( Fig 2D and S2 Fig ) . Therefore , CqsS production increases during QS whereas LuxQ production does not . An obvious mechanism that could underpin the increase in CqsS production that occurs with increasing V . cholerae cell density is regulation of CqsS by QS . To investigate this possibility , we quantified CqsS::3XFLAG levels in V . cholerae QS mutants that are locked at LCD and HCD . As mentioned above ( and see Fig 1 ) , LuxO functions as the QS signal integrator that is phosphorylated and activated at LCD . LuxO D61E is a phosphomimetic allele that constitutively activates expression of the qrr genes . Thus , LuxO D61E confers the LCD state irrespective of cell density due to Qrr-mediated activation of aphA and repression of hapR . Likewise , the ΔhapR strain is locked in LCD mode . Conversely , the LuxO D61A allele cannot be phosphorylated so it is inactive and incapable of activating expression of the qrr genes . Thus , LuxO D61A confers the constitutive HCD state and HapR is produced at all cell densities . Fig 2E and S3A Fig show that , at LCD , the level of CqsS in the LuxO D61E and the ΔhapR strains is equal to that in the WT whereas the LuxO D61A single and the LuxO D61A ΔhapR double mutants possess increased CqsS relative to WT and the locked LCD mutants . Thus , CqsS is upregulated in the absence of LuxO~P and in the absence of HapR . This result shows that regulation cannot occur through HapR , but rather , control must be a consequence of LuxO~P-mediated repression at LCD . We deleted aphA and there was no effect on CqsS production ( S3B and S3C Fig ) . Thus , regulation must occur via the Qrr sRNAs [32–35] . We note that in the HCD-locked mutants , CqsS levels are upregulated 1 . 5-fold relative to WT , while CqsS levels in the WT increase ~4-fold as cells transition from LCD to HCD . We suspect that accumulation of autoinducer over time , which occurs in the WT but not in the locked mutants , is required for full regulation of CqsS production . We examined whether Qrr regulation of CqsS is direct or indirect using a CqsS-mKATE2 translational fusion in E . coli expressing inducible qrr4 . We reasoned that , if CqsS production is controlled by Qrr4 , repression of CqsS-mKATE2 should occur when Qrr4 is induced . We verified that our system works by showing that inducible Qrr4 is capable of repressing a previously-identified direct target , VCA0107-GFP [24] . However , no repression of CqsS-mKATE2 occurred following induction of Qrr4 ( S4 Fig ) . This result , coupled with the fact that there is no obvious base-pairing region between the Qrr sRNAs and the CqsS mRNA , suggests that Qrr regulation of CqsS is indirect . Our next goal was to define the concentration of CAI-1 that induces the CqsS-directed QS response to understand whether QS regulation of CqsS levels influences CAI-1 detection . It is not possible to directly measure CAI-1-CqsS interactions because , like many multi-pass trans-membrane proteins , CqsS has remained recalcitrant to purification and traditional biochemical analyses [31] . We reasoned that knowing the number of CqsS dimers/cell , the concentration of CAI-1 present in cell-free culture fluids , and the timing of QS induction would enable us to infer the in vivo CAI-1:CqsS ratio required for QS activation . Our above data defined the number of CqsS dimers/cell ( Fig 2C ) . We quantified CAI-1 by comparing induction of V . cholerae QS in response to cell-free culture fluids prepared from WT V . cholerae to induction in response to known quantities of synthetic CAI-1 added to cell-free culture fluids prepared from a ΔcqsA ( CAI-1 synthase ) mutant ( S5 Fig ) . We could not accurately quantify CAI-1 by mass spectrometry because , as has been reported previously , hydroxyketones are extraordinarily difficult to differentiate from fatty acids in organically-extracted cell-free culture fluids [36] . We assayed the preparations on a V . cholerae ΔcqsA ΔluxQ mutant strain carrying luciferase . This strain makes no CAI-1 and it lacks the ability to respond to AI-2 . Therefore , it activates light production exclusively in response to exogenously supplied CAI-1 . We call this strain the V . cholerae CAI-1 reporter strain . Using this strategy , we found that the concentration of CAI-1 in V . cholerae cell-free culture fluids increases from 27 nM at OD600 = 0 . 5 to 220 nM at OD600 = 2 . 0 ( Fig 3A ) . CAI-1 was undetectable in our assay at OD600 < 0 . 5 . Together , the CqsS and CAI-1 quantitation allow us to estimate the CqsS:CAI-1 ratios at different cell densities . CqsS levels of ~40 dimers per cell at OD600 = 0 . 2 and ~170 dimers per cell OD600 = 2 . 0 can be converted to 11 pM and 480 pM at those two cell densities , respectively . Therefore , during the transition from LCD to HCD ( i . e . , above OD600 = 0 . 7 ) , CAI-1 is always in at least a 450-fold molar excess relative to the CqsS receptor . The cqsA gene is located adjacent to but in the opposite orientation of the cqsS gene on the V . cholerae chromosome . Thus , their regulation is not obviously linked . Nonetheless , the increase in CAI-1 over the course of growth could be a consequence of QS control of cqsA . To test this possibility , cell-free culture fluids were harvested at LCD from the strains in Fig 2E . Relative CAI-1 concentrations were determined by supplying them to the V . cholerae CAI-1 reporter strain and comparing bioluminescence output . The WT and the LuxO D61E mutant had the same level of CAI-1 in their culture fluids ( Fig 3B ) . The LuxO D61A mutant produced ~7-fold more . Deletion of hapR in the LuxO D61A background diminished induction compared to the single LuxO D61A mutant . Nonetheless , the ΔhapR single and the LuxO D61A ΔhapR double mutant strain still produced twice as much CAI-1 as the WT suggesting that the increase in CAI-1 cannot stem fully from the loss of LuxO~P mediated repression of qrr expression alone . Rather , the results suggest that CAI-1 production is both repressed by LuxO~P at LCD , presumably via the Qrr sRNAs , and activated by HapR at HCD ( Fig 3B ) . To determine the concentration of CAI-1 required to induce QS in vivo , that is , in the presence of a functional LuxPQ pathway , we supplied synthetic CAI-1 to a V . cholerae ΔcqsA strain carrying luciferase . Concentrations of CAI-1 above 200 nM fully induced QS , mimicking what occurs in WT V . cholerae at HCD in response to endogenously-produced CAI-1 ( Fig 3C ) . Induction of bioluminescence occurred at a CAI-1 concentration of 50 nM , corresponding roughly to the concentration of CAI-1 present in WT V . cholerae cell-free culture fluids at which the QS response naturally initiates ( compare data in Fig 3A and 3C ) . Concentrations of CAI-1 below 50 nM failed to activate a QS response ( Fig 3C ) . These results agree with the reported 35 nM Kd of CAI-1 for CqsS [31] . We conclude that the CAI-1 ligand and the CqsS receptor , at least under the conditions we tested , exist at levels within the range that ensures high sensitivity to changing CAI-1 concentrations ( S6 Fig ) . We wondered what consequence feedback on CqsS and CAI-1 has on V . cholerae QS behavior . To assess this , we measured QS induction in a V . cholerae ΔcqsA ΔluxS autoinducer synthase mutant in the presence of one or both autoinducers using bioluminescence as the readout . The double ΔcqsA ΔluxS autoinducer synthase mutant makes 100 , 000-fold less light than WT V . cholerae following overnight growth ( Fig 4A , black vs . white symbols and see S7A Fig ) . Addition of saturating AI-2 increased light production 100-fold at HCD ( blue ) , whereas addition of saturating CAI-1 induced WT levels of light ( i . e . , 100 , 000-fold induction; red ) . Simultaneous addition of both CAI-1 and AI-2 induced light production earlier than did addition of CAI-1 alone ( purple ) . Thus , CAI-1 plays the dominant role in driving QS induction , however , together , the two autoinducers have a synergistic effect . We interpret these results to mean that , in the absence of CAI-1 , the phosphatase activity of AI-2-stimulated LuxQ cannot override the kinase activity of the unliganded CqsS receptor . By contrast , in the absence of AI-2 , the phosphatase activity of CAI-1-bound CqsS can overcome the kinase activity of the unliganded LuxQ receptor . The presence of LuxQ kinase activity , however , delays induction of the QS response to CAI-1 when it is the only autoinducer present . Quantitation of CqsS levels in the above strains shows that the amount of CqsS protein in the ΔcqsA ΔluxS mutant was significantly lower than that in WT at both LCD and HCD ( Fig 4B ) . Addition of either AI-2 ( blue ) or CAI-1 ( red ) increased CqsS protein levels to nearly that present in the WT at LCD and HCD . Simultaneous administration of both CAI-1 and AI-2 ( purple ) increased the CqsS levels to above WT levels , again both at LCD and HCD . These data show that autoinducers control CqsS production via feedback and they act synergistically to do so . Moreover , these data suggest that , even at LCD , V . cholerae can detect the presence of either single autoinducer , and as a consequence , via feedback , increase CqsS production . The feedback-driven increase in CqsS production that occurs at HCD should confer an increase in CqsS-mediated dephosphorylation of LuxO~P relative to that mediated by LuxQ . To examine the relative phosphatase activities of the QS receptors , we measured expression of qrr4 , the direct target of LuxO~P ( S8 Fig ) . For this analysis , we employed a qrr4-luxCDABE transcriptional fusion that is activated by LuxO~P at LCD . qrr4-luxCDABE expression was 50-fold higher in the locked LCD ΔcqsA ΔluxS double autoinducer synthase mutant than in WT V . cholerae because LuxO is maximally phosphorylated in the double synthase mutant ( both CqsS and LuxQ are kinases ) and maximally dephosphorylated in the WT ( both CqsS and LuxQ are phosphatases ) . Addition of saturating AI-2 and addition of saturating CAI-1 to the ΔcqsA ΔluxS double autoinducer synthase mutant reduced qrr4-luxCDABE expression 2-fold and 20-fold , respectively . Importantly , CAI-1 had a 10-fold greater effect than did AI-2 showing that CqsS , the receptor that experiences positive feedback , has the stronger phosphatase activity . Addition of both autoinducers together further repressed qrr4-luxCDABE expression to WT levels , demonstrating the synergy between the two ligands ( S8 Fig ) . We suggest that CqsS exhibits stronger phosphatase activity on LuxO~P than does LuxQ because QS-mediated positive feedback increases CqsS levels , making it the dominant receptor . This hypothesis predicts that altering receptor ratios could transfer dominance to LuxQ . To explore this prediction , we measured the QS output when chromosomal cqsS was placed under the control of the luxPQ promoter . This arrangement reduced CqsS production by 80% compared to when cqsS was expressed from its native promoter ( S9A Fig ) . Under this condition , AI-2 caused premature induction of QS , and the AI-2 input-output range was expanded relative to the AI-2 response shown in Fig 4B ( S9B Fig ) . By contrast , CAI-1 was incapable of inducing the premature response ( S9B Fig ) . Together , these results highlight the importance of receptor ratios in the proper control of QS behavior . Furthermore , the data show that , in WT V . cholerae , positive feedback on CqsS dampens the response to AI-2 . The differential QS induction that V . cholerae exhibits in response to CAI-1 versus AI-2 ( see Fig 4A ) stems from the upregulation of the CAI-1-CqsS pathway that occurs during the QS transition . Specifically , at LCD , there are roughly equal numbers of CqsS and LuxQ dimers ( CqsS:LuxQ = 0 . 7:1 ) . Following the transition to HCD , the CqsS:LuxQ ratio increases to 2 . 5:1 ( Fig 2D ) . We have shown that feedback onto CqsS enables it to disproportionately dephosphorylate LuxO~P ( S8 Fig ) , which , consequently , must increase HapR levels . If so , QS-controlled genes could be regulated exclusively by the CAI-1-CqsS pathway . To investigate this notion , we measured expression of the QS-activated gene hapA [22 , 37] . hapA encodes a protease that is reported to be crucial for V . cholerae dissemination following infection [38 , 39] . We used qRT-PCR to measure hapA transcript in HCD cultures of WT V . cholerae , the ΔhapR mutant , and the ΔcqsA ΔluxS double autoinducer synthase mutant ( Fig 5A ) . HapR activates hapA expression , and consistent with this , the hapA transcript was nearly undetectable in the ΔhapR strain ( black ) . Likewise , the locked LCD double autoinducer synthase ( ΔcqsA ΔluxS , white ) mutant in which HapR is not made ( see Fig 1 ) produced levels of hapA transcript similar to the ΔhapR mutant . Addition of AI-2 to the ΔcqsA ΔluxS strain increased hapA expression 4-fold ( blue ) while addition of CAI-1 increased hapA transcript levels over 50-fold ( red ) . Simultaneous addition of both autoinducers ( purple ) increased hapA expression an additional 2-fold above that following addition of CAI-1 alone . The above experiment , together with the results in S9B Fig , suggests that the stoichiometry of QS receptors dictates the level of hapA expression . To verify this hypothesis , we exchanged QS control of hapA from CqsS to LuxQ by deleting the genes encoding all of the QS receptors except LuxQ ( ΔcqsS ΔcqsR ΔvpsS ) . We quantified the levels of hapA in this strain , which we call the LuxQ+ AI-2+ strain and compared them to a LuxQ+ AI-2- strain ( ΔcqsS ΔcqsR ΔvpsS ΔluxS ) that possesses LuxQ but cannot make AI-2 ( Fig 5B ) . The LuxQ+ AI-2- strain had 44-fold less hapA transcript than the LuxQ+ AI-2+ strain . Addition of AI-2 to LuxQ+ AI-2- strain increased hapA expression 27-fold . Thus , AI-2 can control hapA , but only in the absence of CqsS . We have used hapA as our representative test case . We reason that other QS-controlled genes behave similarly [9 , 40–44] . This result suggests that QS control of receptor ratios dictates discrete gene expression patterns in V . cholerae at HCD , and the downstream effect is dominated by CAI-1-CqsS . The above findings suggest that agonizing the major QS receptor CqsS could influence V . cholerae dispersal , making CqsS an excellent target for small molecule manipulation . V . cholerae CqsS responds to the cognate CAI-1 ligand ( S ) -3-hydroxytridecan-4-one and the close analogs enamino-CAI-1 ( Z ) -3-aminotridecan-2-en-4-one and enamino-C8-CAI-1 , made by related vibrios [30] . However , CqsS does not activate QS in response to CAI-1 analogs with enlarged head groups or shortened acyl tails [31] . We probed the limits of CqsS ligand detection by assessing a library of synthetic compounds for CqsS agonism . We screened 352 , 083 compounds at 20 μM for those that induced bioluminescence 10 , 000-fold in the double synthase mutant ( ΔcqsA ΔluxS ) carrying luciferase . Putative hits were re-tested at a variety of concentrations and active compounds were selected . We employed a secondary screen using the ΔcqsS mutant carrying luciferase to identify the subset of active compounds that require CqsS for function [27] . Nine compounds were identified that met these criteria . They were further subdivided into two groups ( denoted 1A and 1B ) based on their structures ( Fig 6A and 6D ) . None of the compounds share structural similarity with CAI-1 . Class 1A compounds are 5-aminotetrazole derivatives and all are more potent than CAI-1 ( EC50 = 38 nM; Fig 6B ) and all are equally efficacious as CAI-1 at agonizing CqsS ( Fig 6A–6C ) . The most potent compound in this class , compound #1 , is ~10-fold more potent than CAI-1 with an EC50 = 4 nM ( Fig 6C ) . Compound #3 , in which the R2 ethyl group of compound #1 is replaced by the longer propyl side chain , is approximately 4-fold less potent . Compound #6 , containing a free N-H at R2 , loses all activity ( S10 Fig ) , indicating an absolute requirement for a substituent at the R2 position . This result could be due to either stereoelectronic or conformational considerations; the limited dataset does not allow a deeper explanation . The R1 substituent has a minor effect on potency . The N , N-diethylanilino moiety appears to slightly decrease potency relative to the pyrrolodino moiety ( compound #2 vs . compound #3 ) while the isopropyl group shows both a modest increase ( compound #4 vs . compound #3 ) and a slight reduction ( compound #5 vs . compound #1 ) in activity relative to the pyrrolodino moiety , depending on the nature of the R2 substituent . These results suggest that the aniline nitrogen present in compounds #1–3 is not crucial for activity . Class 1B molecules are biphenyl amide derivatives , and all are less potent than CAI-1 ( Fig 6D–6F ) . This initial dataset suggests that cyclic R2 substituents on the secondary amide are preferred over acyclic ones ( compare compounds #8 and #10 vs . compound #7 ) . In addition , substitution of a 2-pyridine ring in place of the benzene ring at position R1 ( compound #9 ) significantly diminished potency . CAI-1 inhibits CqsS autokinase activity , whereas CqsS phosphatase activity is not altered by CAI-1 binding [17] . Thus , in the unliganded state , CqsS kinase prevails , and in the CAI-1-bound state , CqsS phosphatase activity dominates . We tested whether the newly identified synthetic agonists functioned by the same mechanism . CqsS autophosphorylation can be assayed using inverted membrane vesicles containing CqsS protein and measurements of incorporation of 32P . Using a CqsS mutant that cannot transfer the phosphate to the receiver domain ( CqsS D618N ) , we can test the ability of the compounds to specifically inhibit CqsS H194 phosphorylation . Similar to CAI-1 , all of the synthetic agonists inhibited the first biochemical reaction in signal relay: CqsS autophosphorylation ( Fig 7A shows two representative examples , one from Class 1A and one from Class 1B ) . Both classes of synthetic agonists , irrespective of whether their EC50’s are higher or lower than that of CAI-1 , contain compounds with higher efficacy than CAI-1 . Specifically , Fig 7B shows that saturating CAI-1 inhibited 80% of WT CqsS autophosphorylation while compound #10 inhibited 90% and compound #1 inhibited 99% of the activity . The finding that the synthetic compounds function by the same mechanism as CAI-1 to control CqsS signal transduction was surprising given their varied structures relative to the CAI-1 ligand . We therefore examined if the synthetic compounds required the putative CAI-1 binding site for activity . The CqsS F162A , CqsS F166L , and CqsS C170Y variants do not respond to CAI-1 [31] . These amino acid residues are presumed to directly interact with the CAI-1 ligand because changes in the CAI-1 structure that render it inactive can be compensated by companion changes in these key amino acid residues . Specifically , if a bulky moiety is exchanged for a small entity on the CAI-1 ligand , replacing a small amino acid with a bulky one in the CqsS ligand-binding site restores activity , and vice versa . This previous chemical-genetics analysis led to the understanding that in CqsS , F162 and F166 specify the CAI-1 head group and C170 determines the CAI-1 tail length [31] . We used this set of CqsS mutants to explore the requirements for synthetic agonist activity . To do this , we assayed CqsS-directed repression of Qrr4-GFP production . As a reminder , at LCD , when CqsS is unliganded , phospho-relay to LuxO activates Qrr production . At HCD , CAI-1 binding to CqsS terminates Qrr production ( see Fig 1 and S8 Fig ) . Thus , reduced Qrr4-GFP production corresponds to CqsS agonism via inhibition of autokinase activity . CqsS F162A and CqsS C170Y could not be agonized by either class of synthetic compounds discovered here: Qrr4-GFP production remained constant and at high levels even when the compounds were provided at 1 μM ( Fig 7C ) . Therefore , these residues are crucial for synthetic compound detection . Conversely , CqsS F166L , while nearly impervious to CAI-1 , repressed Qrr4-GFP production in response to both Class 1A and 1B compounds , suggesting that F166 is dispensable for signaling by these particular synthetic compounds ( Fig 7C ) . We conclude that the synthetic agonists suppress CqsS kinase activity by employing the native ligand-binding site . The synthetic agonists may make other interactions with CqsS , distinct from those made by CAI-1 , that enable potent binding while eliminating the requirement for the F166 residue .
V . cholerae integrates the information contained in the intra-genus autoinducer CAI-1 and the inter-species autoinducer AI-2 into a shared QS pathway via modulation of the kinase:phosphatase activities of the CqsS and LuxQ receptors , respectively [9] . Deducing the ratio of CAI-1:AI-2 could inform V . cholerae about whether it is the majority or minority species in the environment . We presume that it would be beneficial for V . cholerae to exhibit distinct behaviors depending on whether or not it predominates in any particular consortium . Here , we show that a QS-mediated positive feedback loop promotes increased production of CqsS , but not LuxQ during the QS transition from LCD to HCD . CqsS increases from ~40 dimers per cell at LCD ( OD600 = 0 . 2 ) to ~170 dimers per cell at HCD ( OD600 = 2 . 0 ) , which translates to a change from 11 pM to 480 pM in the culture . We find that CAI-1 production is similarly upregulated , with levels increasing from 27 nM to 220 nM from LCD ( OD600 = 0 . 5 ) to HCD ( OD600 = 2 . 0 ) . Concentrations of CAI-1 that highly exceed those of the CqsS receptor during the transition from LCD to HCD make the increase in CqsS levels inconsequential with respect to the ability of the receptor to detect CAI-1 . A previous study examining yeast strains containing different so-called secrete-and-sense circuits concluded that cells exhibiting high receptor levels relative to ligand levels were “asocial” because they capture their own ligand , which prevents neighboring cells from detecting released ligand [45] . By contrast , cells with low receptor levels coupled with high ligand secretion rates favored social communication by preventing “self-communication . ” Such logic could apply to the V . cholerae QS CqsS receptor . Our results suggest that the crucial consequence of positive feedback on CqsS is to alter the potency of CAI-1-driven intra-cellular signaling relative to that driven by AI-2 , which enables CAI-1-specific QS outputs . Similar to what we show here for hapA , CAI-1-specific regulation in V . cholerae has been reported previously for the heterologous bioluminescence readout , for biofilm formation , and for chemotaxis [9 , 43 , 44] . Here , we have defined the mechanism: asymmetric positive feedback onto CqsS underlies these observations . This feature of the QS architecture could be applicable to other genes in the QS pathway and to other signal transduction systems responsive to multiple inputs in which information is funneled internally through parallel pathways . Receptor ratio modulation should have specific consequences for how effectively V . cholerae monitors its cell density in mono-culture and in mixed-species communities . We take the case of mono-culture first: at LCD , V . cholerae possesses low , but approximately equal numbers of CqsS and LuxQ dimers which fosters roughly equal sensitivity to the CAI-1 and AI-2 signal inputs . Under this condition , exogenous addition of either single autoinducer fails to prematurely induce any QS output whereas simultaneous addition of CAI-1 and AI-2 launches the QS response ( Fig 4A ) . We attribute the lack of response to either individual autoinducer to the overriding kinase activity of the remaining , unliganded receptor . Such an arrangement could protect the system from transient fluctuations in a single autoinducer . Insulation of the QS response from premature activation due to information flowing through a single channel has been reported previously [11 , 46] and could be a conserved attribute of vibrio QS circuits , which , as far as is known , all have similar network architectures . Consistent with this idea , we found that overexpression of cqsS in a ΔcqsA strain resulted in a severe growth defect when exogenous CAI-1 was supplied . We suspect that it is crucial to maintain low CqsS levels at LCD because the mis-regulation of a downstream QS target ( s ) under this condition is lethal . We are currently investigating the mechanism underpinning the LCD growth defect . With respect to receptor ratios at HCD in V . cholerae mono-cultures , here we find that a mechanism exists to ensure an asymmetric increase in production of CqsS relative to LuxQ during the transition into QS mode . Previously , receptor ratio modulation was studied in the related bacterium , V . harveyi . V . harveyi employs three autoinducer-receptor pairs and possesses a feedback loop controlling receptor levels for the AI-1-LuxN pathway , a pathway that does not exist in V . cholerae [47 , 48] . Positive feedback on LuxN increased the AI-1 input range , allowing V . harveyi to enhance its sensitivity to AI-1 relative to AI-2 ( and presumably CAI-1 , but that was not explicitly tested ) at HCD [49 , 50] . Our results support this notion: positive feedback on CqsS in V . cholerae likewise diminishes the relative sensitivity to AI-2 at HCD . Here we find that this feature of the network endows V . cholerae with the ability to control genes exclusively in response to one particular autoinducer , CAI-1 , and indeed , CAI-1 alone is capable of fully launching the V . cholerae QS response ( Figs 4 and 5 ) . This finding differs fundamentally from what was shown in V . harveyi: the V . harveyi QS response cannot be fully induced unless all autoinducers are present [40 , 46] . We know that , in the absence of the CqsS receptor , AI-2 signaling through LuxQ can fully induce the V . cholerae QS response ( Fig 5B ) . Thus , the asymmetry in CAI-1 response that we discovered stems exclusively from positive feedback on CqsS which changes the stoichiometry of the two major QS receptors [49–53] . Biasing the system to favor the intra-genus CAI-1 autoinducer at HCD could be a type of kin-discrimination , ensuring that expensive public goods are not produced until sufficient numbers of kin are present [54] . Now we consider the consequences of feedback on CqsS for V . cholerae in multi-species consortia . The V . cholerae lifecycle involves transitions between the human host and the aquatic environment [55–57] . We first consider the case in the environment . If V . cholerae is at LCD and other vibrios are present , V . cholerae should rapidly launch its QS cascade because other vibrios produce AI-2 and different CAI-1 moieties . V . cholerae CqsS responds to CAI-1 , enamino-CAI-1 ( Ea-CAI-1 ) , and Ea-C8-CAI-1 . Common marine vibrio species produce these molecules [30] . Thus , if other vibrio species are present , both autoinducer signals could be present , which is the necessary condition to enable rapid activation of the V . cholerae QS system . Presumably , vibrios that use the same autoinducer and share a niche could more successfully colonize habitats via coordinated production of public goods . We predict that this feature could be useful in enabling V . cholerae , together with its close vibrio relatives , to efficiently colonize marine hosts such as copepods , algae , and fish [56 , 58–61] . We consider the scenario in which V . cholerae enters a multi-species environment devoid of vibrios: the human host . In this case , we expect AI-2 to be the predominant signal encountered since AI-2 is broadly made by bacteria [29 , 62–64] . V . cholerae should not react to AI-2 because , as we have shown , at LCD , the V . cholerae QS system is impervious to any single autoinducer . Under this condition , as V . cholerae grows within the consortium , the CAI-1 concentration should track with increasing V . cholerae cell density since , presumably , only V . cholerae is contributing CAI-1 to the human host milieu . By contrast , AI-2 should accumulate disproportionately relative to V . cholerae cell numbers . As V . cholerae reaches HCD , biasing the response to CAI-1 via feedback on CqsS would inflate the ability of V . cholerae to detect numbers of its kin within a mixed-species community . Indeed , feedback on the V . harveyi AI-1-LuxN pathway was proposed to allow V . harveyi to “pay more attention” to AI-2 at LCD and to AI-1 at HCD [49 , 50] . Our results with V . cholerae are similar with respect to CAI-1 ( Fig 4A ) . Now we put the above results into the context of infection . V . cholerae presumably enters the host in the LCD QS state and so it is in biofilm-formation-mode and actively producing virulence factors [9 , 22 , 43] ( see Fig 1 ) . We presume this genetic program primes V . cholerae for successful infection . Upon entrance into the small intestine , V . cholerae likely encounters a bolus dose of AI-2 but that does not affect biofilm formation or virulence factor expression since both CAI-1 and AI-2 are required at LCD to alter the QS response . We suggest that this arrangement allows V . cholerae to delay launching its dispersal program until it has grown to significant cell numbers in the host [65] . Positive feedback onto CqsS , however , ensures that once launched , V . cholerae remains committed to the dispersal program by exaggerating its response to CAI-1 irrespective of whether AI-2 levels are high or low or fluctuating . We do note that high AI-2 levels would intensify the expression of the V . cholerae HCD program . Importantly , accumulation of the threshold CAI-1 concentration required for the QS response would be the key event that signifies the completion of the infection cycle since CAI-1 alone is sufficient to cause a QS response while AI-2 cannot trip the system in the absence of CAI-1 . The CqsS-dominated QS system of V . cholerae makes CqsS a possible target for small molecule manipulation . Here , we have identified two classes of CqsS synthetic agonists . While they are structurally dissimilar from the native CAI-1 autoinducer , they inhibit CqsS H194 autophosphorylation similarly to CAI-1 . Interestingly , some of the synthetic agonists can inhibit CqsS autophosphorylation more effectively than can the native CAI-1 autoinducer . Conceivably , such a CqsS agonist could more potently and more prematurely induce QS than does CAI-1 resulting in the execution of the V . cholerae host-dispersal program . Thus , small molecule CqsS agonists could be envisioned as treatments or preventatives for cholera disease , especially in the presence of microbiota-produced AI-2 . Collectively , the results presented in this work support the growing hypothesis that bacteria have mechanisms to uncouple some behaviors from the bulk of the QS-controlled gene expression program , and in so doing , uncouple a portion of the QS response from the constraint of strict cell-density dependence . Pseudomonas aeruginosa uses two autoinducers with distinct decay rates to accomplish this feat [66] . The alternative VqmR/VqmA QS pathway in V . cholerae responds to a microbiota-produced cue called DPO to control biofilm production independently of the genes controlled by the canonical QS signal transduction pathway discussed here [67] . In addition , receptor ratio regulation could provide a mechanism for V . cholerae to distinguish when it is a majority species in a consortium and , in response , to control a subset of QS target genes . We expect that such information-processing features embedded in larger QS circuits provide plasticity to the systems enabling bacteria to more successfully and appropriately orchestrate group behaviors .
All V . cholerae strains are derived from wild-type C6706str2 [68] . E . coli S17-1 λpir was used for cloning and E . coli C43 ( DE3 ) was used for protein overexpression . V . cholerae and E . coli were grown in LB medium at 37°C with shaking . Bioluminescence assays were conducted in SOC medium supplemented with tetracycline . Antibiotic concentrations used are as follows: ampicillin , 100 mg/L; kanamycin , 100 mg/L; chloramphenicol , 10 mg/L; tetracycline , 10 mg/L; streptomycin , 5 g/L; polymixin B , 50 U/L . Chemical syntheses of CAI-1 and AI-2 have been described [3 , 31 , 69 , 70] . To identify CqsS agonists , the Broad Institute’s 352 , 083 compound library was screened as described [27 , 71] . Alterations to the V . cholerae genome were generated using the pKAS32 allelic exchange method [72] . Construction of 3XFLAG fusions on pKAS32 plasmids was accomplished via splicing by overlap extension at the C-termini . In all strains carrying CqsS::3XFLAG in which cqsA was also mutated ( AH366 , AH367 , AH370 , AH371 , and AH468 ) , the cqsS gene was first fused to 3XFLAG on the chromosome . cqsA was next disrupted by deleting a T ( TTT to AA- ) in codon 9 , the consequence of which , was introduction of a stop codon at codon 14 . These strains are used in Fig 3 , Fig 4 , S7 Fig and S9 Fig . In those figure panels , for simplicity , we call the cqsA mutation ΔcqsA . The CqsS::3XFLAG overexpression vector was constructed by overlap extension PCR from the pET21b plasmid containing the gene encoding CqsS::6XHis [17] . The LuxQ::3XFLAG overexpression vector was constructed by amplifying luxQ from the genome of a V . cholerae strain ( AH420 ) carrying luxQ::3XFLAG by exploiting flanking 5’ NcoI and 3’ BamHI sites . The insert was ligated into similarly digested pET21b . The NcoI site introduced a mutation ( D2N ) into the gene . It was corrected with using Quikchange II XL Site-Directed Mutagenesis ( Stratagene ) . PCR reactions used iProof DNA polymerase ( Bio-Rad ) . To quantify the function of the CqsS::3XFLAG construct in V . cholerae , overnight cultures of V . cholerae strains carrying luxCDABE were diluted to OD600 = 0 . 005 in fresh medium and grown at 30°C with shaking . Every hour , bioluminescence and OD600 were measured on a Tri-Carb 2810 TR scintillation counter and DU800 spectrophotometer , respectively . Samples with OD600 > 0 . 7 were back-diluted 10-fold prior to analysis to ensure measurements could be made within the linear range of the spectrophotometer . CAI-1 was dissolved in DMSO and AI-2 was dissolved in water . The autoinducers were added at the indicated concentrations at the time of dilution . DMSO was used as the negative control and 100 μM of boric acid was added to cultures containing AI-2 . For relative CAI-1 activity assessment , an overnight culture of the V . cholerae CAI-1 reporter strain ( WN1102: ΔcqsA , ΔluxQ/pBB1 ) was diluted 1:10 into sterile medium and 30% ( v/v ) cell-free culture fluids prepared from the indicated V . cholerae strains were added to the diluted reporter strain . The plasmid pBB1 carries the V . harvyei luxCDABE genes . Bioluminescence and OD600 were measured after the cultures were incubated at 30°C for 2 . 5 h with shaking . To quantify absolute CAI-1 concentration , the bioluminescence from the CAI-1 reporter strain supplemented with 30% cell-free culture fluids from WT V . cholerae was compared to the bioluminescence from the reporter strain supplemented with known concentrations of synthetic CAI-1 in 30% cell-free ΔcqsA V . cholerae cell-free culture fluids . DMSO was added to the cell-free WT culture fluids as a negative control . The concentration of endogenously produced CAI-1 in the cell-free culture fluids was extrapolated from the log ( agonist ) vs . response variable slope calculation for the synthetic CAI-1-directed bioluminescence output using Prism software . The relative phosphatase activities of the QS receptors were measured in WT V . cholerae and in the ΔcqsA ΔluxS double autoinducer synthase mutant using a qrr4-luxCDABE transcriptional fusion . Overnight cultures were diluted 1:20 in fresh medium in the presence and absence of exogenous autoinducers and incubated 4 h at 30°C with shaking . Bioluminescence was measured on a Tri-Carb 2810 TR scintillation counter . Bioluminescence was normalized to OD600 , which was measured on a DU800 spectrophotometer . To determine the EC50 of synthetic CqsS agonists , the CAI-1 reporter strain WN1102 was diluted 1:20 in fresh medium . The indicated concentrations of CAI-1 or synthetic agonist were added in triplicate to 96-well plates and incubated for 4 h at 30°C with shaking . Bioluminescence and OD600 were measured on an Envision 2103 Multilabel Reader ( Perkin Elmer ) . We note that the baseline for bioluminescence is lower on the plate reader compared to that from the scintillation counter . We use relative light units rather than absolute light output , which circumvents issues arising from the different sensitivities of the instruments . The EC50 of each compound was calculated using Prism software . To assess the ability of synthetic agonists to control QS genes , we used a qrr4-gfp construct in an assay that has been described [31] . To assay qrr4 regulation of cqsS , the relative fluorescence of translational fusions ( VCA0107B-GFP and CqsS-mKATE2 ) was measured in E . coli BWR1 . E . coli carrying one of these plasmids also contained either an empty plasmid ( pZA31 ) or pZA31 carrying anhydrous tetracycline-inducible qrr4 . Qrr4-mediated repression of CqsS-mKATE2 was assessed . VCA0107-GFP is a known Qrr4-repressed target and was used as a positive control to show the system was functional . Overnight cultures were diluted 1:50 into fresh medium and aliquotted in triplicate into 96 well plates . Concentrations of anhydrous tetracycline from 0 . 4 to 100 ng/μl were added to induce qrr4 expression and the plates were incubated overnight at 30°C with shaking . OD600 and fluorescence were assessed using an Envision 2103 Multilabel Reader ( Perkin Elmer ) . E . coli C43 ( DE3 ) harboring pET21b plasmids carrying CqsS::6XHis or CqsS::6XHis D618N [17] were grown at 37°C with shaking in LB supplemented with kanamycin . Overnight cultures were diluted 1:100 in fresh LB with kanamycin and grown at 37°C with shaking for 3 h . Protein production was induced by the addition of 300 μM IPTG followed by growth overnight at 18°C with shaking . Cells were harvested by centrifugation at 5 , 000 g for 20 min , resuspended in lysis buffer ( 50 mM Tris pH 8 . 0 , 200 mM NaCl , 5 mM MgCl2 , complete mini EDTA-free protease inhibitor ( Roche , #11836170001 ) ) , and lysed under 15 , 000 psi with a high-pressure pneumatic high shear fluid processor ( Microfluidics , M-110Y ) . Cell lysates were clarified at 9 , 300 g for 30 min and the clarified supernatant was subjected to ultra-centrifugation at 180 , 000 g for 1 h . Membrane pellets were resuspended in kinase buffer ( 50 mM Tris pH 8 . 0 , 100 mM KCl , 5 mM MgCl2 , 10% ( v/v ) glycerol ) and the CqsS proteins were quantified by western blot using CqsS::6XHis protein purified in detergent as the standard ( see below ) . Phosphorylation assays were performed using inverted membrane vesicles containing 10 μM of WT CqsS or CqsS D618N protein as described [17] . To purify CqsS::3XFLAG protein , inverted membrane vesicles were prepared and following the ultra-centrifugation step , membrane pellets were resuspended in membrane extraction buffer ( 20 mM Tris pH 8 , 100 mM NaCl , 20% ( v/v ) glycerol ) and solubilized by rocking with 2% foscholine-12 ( FC12: Avanti , #29557-51-5 ) at 4°C for 2 h . Solubilized membranes were incubated with M2 FLAG resin ( Sigma-Aldrich , A2220 ) at 4°C for 2 h . Solubilized membranes were separated from the resin using gravity flow . The flow-through was passed over the resin twice more , and finally , the resin was washed with 10 column volumes of wash buffer ( 50 mM Tris pH 7 . 4 , 150 mM NaCl , 0 . 2% FC12 ) and the protein was eluted with elution buffer ( 0 . 1 M glycine pH 3 . 5 , 0 . 2% FC12 ) . The eluted fractions were neutralized with 2% ( v/v ) 1 M Tris pH 8 . 0 . Protein concentration was determined by Bradford assay ( ThermoFisher , #23246 ) , purity is shown in S12 Fig . LuxQ::3XFLAG was purified by the identical procedure . CqsS::6XHis protein in detergent was purified as follows: Solubilized membranes were incubated with nickel resin ( Qiagen , #30410 ) at 4°C for 2 h followed by separation from the resin using gravity flow . The flow-through was passed over the gravity column once again , and after that , the resin was washed with 10 column volumes of nickel binding buffer ( 20 mM Tris pH 8 , 20 mM imidazole , 300 mM NaCl , 10% ( v/v ) glyercol , 4 mM 2-mercaptoethanol , 0 . 2% FC12 ) and the protein was eluted with 400 mM imidazole . Imidazole was removed by gel filtration on an S200 column with 20 mM Tris pH 8 buffer containing 150 mM NaCl , 5% ( v/v ) glycerol , 1 mM Tris- ( 2-carboxyethyl ) phosphine ( TCEP ) and 0 . 2% FC12 . Protein concentration was determined by Bradford assay ( ThermoFisher , #23246 ) . The number of CqsS dimers per V . cholerae cell was calculated by quantifying the concentration of CqsS per cell and converting it to dimer number . Cell density was determined by diluting and plating cultures on LB agar in triplicate . Aliquots of WT or mutant V . cholerae strains carrying CqsS::3XFLAG were subjected to centrifugation for 15 min at 8 , 000 g and the resulting pellets were flash frozen . The cells in the pellets were lysed for 10 min at 25°C by resuspending in 100 μl Bug Buster ( Novagen , #70584–4 ) supplemented with 0 . 5% Triton-X , 50 μl/ml lysozyme , 25 U/mL benzonase , and 1 mM phenylmethylsulfonyl fluoride ( PMSF ) per 1 . 0 OD600 of pelleted culture . Protein from the cell lysate was solubilized with SDS-PAGE buffer for 1 h at 37°C . Various amounts of the samples were loaded onto 4–20% Mini-Protein TGX gels ( Bio-Rad , #456–1096 ) alongside known amounts ( 0 . 06–2 ng ) of detergent-purified CqsS::3XFLAG for comparison . The samples were electrophoresed for 1 . 5 h at 100 v . Proteins were transferred from the gels to PVDF membranes ( Bio-Rad , #162–0174 ) for 1 h at 4°C at 100 v in 25 mM Tris buffer , 190 mM glycine , 20% methanol . Membranes were blocked for 1 h in 5% milk , washed 3 times with TBST ( 140 mM NaCl , 20 mM Tris-HCl pH 7 . 6 , 0 . 1% Tween ) and incubated for 1 h with 0 . 2 μg/ml monoclonal Anti-FLAG-Peroxidase antibody ( Sigma , A8592 ) in TBST with 3% BSA ( Sigma , A3059 ) . After washing three times with TBST , membranes were exposed using the Amersham ECL western blotting detection reagent ( GE Healthcare , RPN2106 ) for 40 s . Linear regression analysis of band intensity ( volume ) of detergent-purified CqsS::3XFLAG was performed for each gel to calculate the amount of CqsS in the band from the V . cholerae samples . Protein amount was converted to molarity using the MW of the CqsS::3XFLAG construct of 83 . 61 kDa . The CqsS concentration and the cell density of each V . cholerae sample were used to estimate CqsS dimers per cell . V . cholerae overnight cultures were back-diluted 1:100 into fresh medium and grown to OD600 = 2 . 0 . Autoinducers were added at the indicated concentrations at the time of dilution . RNA was harvested from three independent cultures using the RNeasy mini kit ( Qiagen , #4104 ) and cDNA was generated as described [73] with SuperScript III reverse transcriptase ( Invitrogen , #18080–044 ) using 1 . 5 μg of RNA . Real-time PCR analyses were performed as described [73] using a QuantStudio 6 Flex Real-Time PCR detection system ( ThermoFisher ) and the Sybr Green mix ( Quanta , #95074 ) . Quadruplicate technical replicates for each experiment were analyzed by a comparative CT method ( Applied Biosystems ) in which the relative amount of hapA [74] was normalized to the internal hfq control RNA to determine the relative RNA levels . hapA transcript levels in V . cholerae mutants were normalized to WT levels . | Bacteria communicate , count their numbers , and act as collectives using a process called quorum sensing . Quorum sensing relies on the production , release and group-wide detection of small molecules called autoinducers . Curiously , quorum-sensing bacteria often employ multiple autoinducers that funnel information into a common signal transduction pathway . This feature of quorum-sensing systems has made it difficult to understand if or how individual autoinducers could produce unique quorum-sensing responses making the benefit of using multiple autoinducers mysterious . The global pathogen Vibrio cholerae uses two , seemingly redundant , quorum-sensing autoinducers to control pathogenicity . Here , we discover a mechanism that drives an increase in production of one autoinducer receptor , but not the other . This asymmetry allows V . cholerae to disproportionately respond to one autoinducer over the other autoinducer , and in so doing , control behaviors exclusively in response to the dominant autoinducer . Our work solves the issues of whether and how V . cholerae achieves autoinducer-specific control of quorum-sensing behaviors . In this manuscript , we also identify nine synthetic compounds that potently mimic the dominant V . cholerae autoinducer . These compounds represent leads for V . cholerae prophylactics and/or therapeutics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"pathogens",
"vibrio",
"gene",
"regulation",
"regulatory",
"proteins",
"microbiology",
"enzymology",
"dna-binding",
"proteins",
"operons",
"electromagnetic",... | 2017 | Asymmetric regulation of quorum-sensing receptors drives autoinducer-specific gene expression programs in Vibrio cholerae |
Stochastic expression of genes produces heterogeneity in clonal populations of bacteria under identical conditions . We analyze and compare the behavior of the inducible lac genetic switch using well-stirred and spatially resolved simulations for Escherichia coli cells modeled under fast and slow-growth conditions . Our new kinetic model describing the switching of the lac operon from one phenotype to the other incorporates parameters obtained from recently published in vivo single-molecule fluorescence experiments along with in vitro rate constants . For the well-stirred system , investigation of the intrinsic noise in the circuit as a function of the inducer concentration and in the presence/absence of the feedback mechanism reveals that the noise peaks near the switching threshold . Applying maximum likelihood estimation , we show that the analytic two-state model of gene expression can be used to extract stochastic rates from the simulation data . The simulations also provide mRNA–protein probability landscapes , which demonstrate that switching is the result of crossing both mRNA and protein thresholds . Using cryoelectron tomography of an E . coli cell and data from proteomics studies , we construct spatial in vivo models of cells and quantify the noise contributions and effects on repressor rebinding due to cell structure and crowding in the cytoplasm . Compared to systems without spatial heterogeneity , the model for the fast-growth cells predicts a slight decrease in the overall noise and an increase in the repressors rebinding rate due to anomalous subdiffusion . The tomograms for E . coli grown under slow-growth conditions identify the positions of the ribosomes and the condensed nucleoid . The smaller slow-growth cells have increased mRNA localization and a larger internal inducer concentration , leading to a significant decrease in the lifetime of the repressor–operator complex and an increase in the frequency of transcriptional bursts .
Transcriptional and translational regulatory networks control the phenotype of modern cells , regulating gene expression in response to changing environmental conditions and/or biological stimuli . It has been well established that intrinsic noise in gene regulation results from the discrete biochemical nature of the process [1] . There is also an extrinsic component to the total noise arising from cell-to-cell variation in the number of copies of the transcription and translation machinery ( transcription factors , RNA polymerases , ribosomes , etc ) [2]–[4] . Stochastic noise can lead to different phenotypic outcomes for a cellular population and , in certain fluctuating environments , the resulting heterogeneous population can be more optimal for growth than would be a population containing a single phenotype [5] , [6] . Theoretical modeling of stochasticity in gene expression has been a topic of intense study in the last decade and has greatly increased our understanding of the effect that statistical noise has on gene regulation ( for reviews see [7]–[11] ) . Without detailed information regarding spatial heterogeneity within a cell , models of stochastic gene expression are typically expressed in terms of the chemical master equation ( CME ) , which describes the time evolution of the probability for a chemical system to be in a given state [12] . Various analytical methods including moment generating functions [1] , [3] , [13] , the Langevin and Fokker-Planck equations [14] , linear noise approximation [4] , and many-body theory [15] are used to study such models of gene expression . Computer simulations , usually based on a variant of Gillespie's stochastic simulation algorithm ( SSA ) [16] are also widely employed to analyze gene network models that are too complex to be amenable to analytical modeling [17] , [18] . Such theoretical studies have predicted and experimental measurements have shown [2] , [19]–[23] that populations of cells can be quite heterogeneous , even when starting from an initially identical state . The large variance in the population distribution is usually ascribed to bursting in the process of gene transcription . Two models have been developed which can be used as a framework for quantitatively analyzing population distributions to infer the underlying gene expression kinetics . The burst model ( Figure 1A ) of Friedman et al . [24] is based on the assumption that an mRNA's lifetime is short compared with that of its protein product . In that case , proteins will be produced in independent bursts with exponentially distributed sizes . The solution to the stationary probability distribution of protein in the continuous CME formulation of the model was shown to be the Gamma distribution where the and parameters were interpreted to be the frequency of transcriptional bursts relative to the protein lifetime and the mean number of proteins produced per burst , respectively . Shahrezaei and Swain [25] further developed the analytical theory of gene expression , by deriving not only the time-dependent probability distribution for the burst model , but also the steady-state distribution for a two-state model of gene expression ( Figure 1B; three-stage model in their nomenclature ) . In the two-state model a gene alternates between transcriptionally active and inactive states with constant rates . Their analytical distributions show that in addition to large variance within a population , bimodality can appear when transitions between the active and inactive states are slow . A similar model has also been used to analyze the switching behavior of a population due to rare large events versus the cumulative effect of many small events [26] . Computational modeling can greatly assist in understanding genetic systems where complexity exceeds the capacity of analytical solutions . In a model for an inducible genetic switch incorporating more of the complexity present in real biological systems ( Figure 1C ) , the transitions between the active and inactive transcriptional states are no longer constant but depend upon an external inducer likely in a nonlinear manner . The positive feedback ( PFB ) loop changes the network topology by introducing an additional regulatory link . Both of these differences provide additional sources of noise in the circuit that may affect the probability distributions . Combining computer modeling of a complete genetic circuit with analysis using simplified analytical models can help to provide an overall picture of the dynamics of such a system . Further complexity in modeling real biological systems comes from the spatial heterogeneity within a cell and molecular crowding in the in vivo environment . It is becoming apparent that the cell is not a well-stirred system [27]–[29] . Studies using cryoelectron tomography techniques [30]–[34] have revealed that individual macromolecules are not necessarily uniformly distributed inside the cell , but may be clustered in a spatially dependent manner . Spatial organization can affect reaction kinetics by increasing local concentrations of reactants and enzymes . Additionally , crowding and non-specific molecular interactions in the in vivo environment can lead to anomalous subdiffusive behavior for macromolecules , as measured experimentally [35] , [36] and by computational modeling of bacterial cytoplasmic environments [37]–[39] . Accounting for spatial heterogeneity is a challenge to computational biology that must eventually be met and several such modeling studies have been undertaken [37]–[44] . Stochastic modeling of gene expression circuits in a three-dimensional bacterial cell poses several difficulties , both computational and informational in nature . Recently a “lattice microbe” method [37] was developed using GPU ( graphics processing unit ) computational accelerators to simulate diffusion of macromolecules within a modeled Escherichia coli cell packed with a distribution of obstacles according to reported proteomics data . It implemented a multiparticle reaction-diffusion algorithm on a three-dimensional lattice to perform simulations of cell-scale systems . With the lattice microbe method one can observe anomalous diffusion of macromolecules and track diffusive-reactive processes over the timescale of the cell cycle , with spatial resolution from 2–16 nm . On the informational side , painstaking efforts must be undertaken to obtain parameters for the models . Kinetic parameters , which are often obtained under in vitro conditions , must be validated by comparing modeling results to published experiments . Recent time-lapse fluorescence microscopy experiments have been able to track dynamic behavior for individual macromolecules in vivo [21] , [45] , providing an additional source for model parameters . Parameters obtained from in vivo single-molecule experiments are uniquely suited for stochastic modeling , as they provide population distributions not simply mean values from ensemble measurements . Equally importantly such parameters are measured under in vivo conditions and incorporate the effects of the cellular environment . Also , super-resolution imaging studies [46]–[49] provide further spatial information to complement the cryoelectron tomography data . We present here a computational study of gene expression noise in the inducible genetic switch shown in Figure 1C using both well-stirred and spatially resolved models . Spatial models of E . coli cells were constructed to approximate cytoplasmic crowding under both rapid and slow growth phenotypes , with the latter being based on data from cryoelectron tomography [50] . Both spatial models were simulated using the lattice microbe method [37] . The genetic switch was based on the well-characterized E . coli lactose utilization system , parameterized using measurements from a recent series of in vivo single-molecule fluorescence studies [21] , [22] , [51] as well as published in vitro rate constants . We report the contributions to intrinsic noise from the regulatory elements of the inducible genetic circuit as well as the extrinsic noise due to in vivo crowding . Using the slow-growth model we investigate the effect of using experimentally determined cellular architecture in reaction-diffusion models , with implications for effects due to cell growth . Comparing the noise from the inducible genetic switch to the bursting and two-state models described above ( Figure 1A , B ) , we consider what improvements in both modeling and experimental efforts are needed to develop stochastic models of gene expression with predictive power regarding phenotype switching and heterogeneity in cellular populations .
Since the stochastic switch model is more complex than can be solved using analytic methods , we used computational Monte Carlo methods to sample the master equation and estimate the probability distributions . Two stochastic approaches were used to simulate the lac kinetic model: a well-stirred method using the CME and a spatially resolved method based on the reaction-diffusion master equation ( RDME ) . The RDME model of the lac circuit can be thought of as a superset of the CME model in that all of the kinetic rates used for modeling reactions in the CME based model are also used in the RDME model , but with additional parameters regarding the spatial localization of particles and their diffusion in three-dimensional space . We analyzed the capability of the burst and two-state analytic models of gene expression to recover parameters from our stochastic simulations of an inducible switch by fitting molecular distributions . We used a maximum likelihood method to estimate the model parameters . Briefly , the likelihood of the model parameters having produced a set of observations is given bywhere is the conditional probability of observation occurring given the parameters . The parameters that maximize this likelihood function are those that describe the best fit of the model to the data , assuming a uniform prior distribution for the parameter probabilities . To find the best parameters for a model of gene expression , was calculated using the model's steady-state probability density function with the values being the protein counts from the 10 , 000 simulations . The parameter values that minimized the negative log of the likelihood function were then found using downhill simplex minimization as implement in the Matlab fminsearch function . We estimated the confidence intervals for different sample sizes by taking 1000 random sets of either 50 or 200 cells from the full set of 10 , 000 and performed maximum likelihood estimation on each of these data sets . The confidence range for each parameter was then defined by the middle 95% of the values obtained during these random resamplings . The burst model was first expressed in terms of parameters and by Friedman et al . [24] as the Gamma distribution . However , since our stochastic simulations produced discrete protein counts , we used the discrete formulation for the steady-state probability density derived by Shahrezaei and Swain [25] in terms of a negative binomial distribution ( 10 ) with parameter being the burst frequency ( bursts per mean protein lifetime ) and being the burst size ( proteins produced per burst ) . The two-state model was fit using the steady-state probability density function derived by Shahrezaei and Swain [25]: ( 11 ) ( 12 ) In this expression the parameters are , , ( the activation rate ) , and ( the inactivation rate ) , the latter two being expressed in units of mean protein lifetime . Additionally , , , , and is Gauss's hypergeometric function . Fitting with all four parameters free often resulted in convergence in a local minima , so we adopted a fitting procedure whereby we first constrained the and parameters and fit only and to obtain initial estimates of these two parameters . In the fully induced state the above probability density function reduces to a negative binomial distribution with no dependence on or , only and . Since neither nor depend on inducer concentration , it is a reasonable approximation to use the values for and in the fully induced state as initial estimates for all inducer concentrations . After obtaining an initial fit for and , we then performed another fit with and unconstrained and with and allowed to vary ±5% . This procedure resulted in convergence at a higher likelihood score than when all four parameter were fit simultaneously for all distributions except one .
In a recent in vivo single-molecule fluorescence study , Choi et al . measured the distributions of a fluorescent reporter protein under control of the lac operator in individual E . coli cells at various inducer ( TMG ) concentrations [22] . They performed the measurements in the absence of LacY's positive feedback by replacing its gene with that of the membrane protein Tsr in the lac operon . This enabled an accurate determination of the protein distribution produced by the circuit at a given inducer concentration without any confounding non-linear effects due to enhancement of the internal inducer concentration by LacY . In the absence of DNA looping , they were able to fit their observed distributions to a gamma distribution , where was interpreted as the frequency of transcriptional bursts relative to the protein lifetime and as the mean number of proteins produced per burst . They observed a relatively constant value for the burst frequency of 3–4 and a linearly increasing relationship between burst size and inducer concentration at low to intermediates concentrations . To understand the origin of the linear relationship between burst size and inducer concentration and to reproduce this behavior in our model , we derived an expression for the burst size as a function of kinetic parameters in our model . As long as bursts are infrequent relative to protein degradation , i . e . once a free operator is bound with a repressor it remains bound for a significant fraction of the cell cycle , transcriptional bursting from the lac operon can be modeled as a Markov process with competition between RNA polymerase ( RNAP ) and the various LacI species for binding to the free operator ( see Figure 4 ) . Transcription initiation by RNAP was modeled as a pseudo first order process ( Equation 4 ) , with a rate constant of . The two repressor states with potentially significant binding affinity were and , shown in Equations 1 & 2 . Free repressor binds with free operator with a rate constant of resulting in a pseudo first order rate of . Given the current debate surrounding the binding affinity of the state to the operator , we set the rate constant to be proportional to the free repressor binding constant and analyzed the effect of varying the proportionality constant on the pseudo first order rate . This model of transcriptional bursting assumes that the binding of to the free operator is negligible at low inducer concentrations by assuming and ignoring Equation 3 . In practice , this condition was satisfied when . We used the upper limit in our model , which is within the range experimentally reported [76] . Following the unbinding of a repressor from the repressor–operator complex , the probability of transcription initiation ( and subsequent mRNA creation ) occurring at the free operator as opposed to a repressor re-binding is ( 13 ) The probability of a given number of consecutive transcription initiation events ( the size of the mRNA burst ) then follows a geometric distribution with of which the mean is . However , repressor unbinding events that produce no mRNA are not observable as a burst , therefore the mean number of mRNA produced in a transcription bursts ( B ) is ( 14 ) Combining Equations 13 and 14 gives the expression for the mean transcription burst size in terms of the rate constants for transcription initiation and repressor binding ( 15 ) Given the inducer mass balances ( see Supporting Text S1 ) and the expression for the total number of repressor dimers , one can derive the equilibrium concentrations of the two repressor specieswhere is the inducer concentration at which half of the repressor monomers are bound to an inducer molecule . Substituting and into Equation 17 gives the expression for the transcription burst size as a function of inducer concentration ( 16 ) From this last equation it is clear that the transcription burst size will be linear over the entire range of inducer concentrations only when . Figure 5 shows the effect of varying , of particular interest are the very low values . When , the transcription burst size does not linearly increase over the range of inducer concentrations for which this behavior has been reported ( 0–200 ) . In the model here formulated , a linear relationship between size and inducer concentration exists only when the binding affinity of for the free operator is comparable to that of . For our simulations , we therefore chose , such that , as this value assumed no effect on the unbound repressor monomer due to a single bound inducer and gave a strictly linear relationship for all inducer concentrations . To obtain values for the model parameters , , and , we used the distributions for LacY reported by Choi et al . [22] , specifically the inferred burst frequency ( bursts per cell cycle ) and size parameters ( and ) from their gamma distribution fits . From Equation 18 , the mean transcription burst size as a function of inducer concentration is . This equation is linear in inducer concentration and by fitting it ( multiplied by the mean number of proteins produced per mRNA ) to the experimental protein burst sizes , as shown in Figure 6 , one can constrain the kinetic parameters . The y-intercept of the line fixes the ratio of transcription to repression in the uninduced state ( ) and the slope can then be used to obtain = 17 . 6 for TMG . The linear fit , however , only fixes the ratio between and . To recover unique values for these two rate constants , we next considered the mean duration of each transcription burst . The interpretation of the shape parameter of the gamma distribution as the burst frequency is only meaningful if the burst duration is short compared to the protein lifetime . In that case , individual exponentially sized bursts can be considered exponentially distributed in time and therefore act independently to give rise to a gamma distribution of protein abundance . In setting rate constants for the model , then , we wanted to ensure that the burst duration was appropriately short . The burst duration is simply the mean time for a repressor to bind to a free operator . Given a constant , a linear relationship between burst size and inducer concentration also implies a linear relationship between and inducer concentration as can be seen from ( 17 ) where in the last step . For TMG , the linear relationship between burst size and inducer concentration extended to at least ∼200 , which is ∼11 times the value for TMG of 17 . 6 . From Figure 7 it can be seen that the interpretation of as the burst frequency begins to break down once is >5% of the protein lifetime . Using 5% of the protein lifetime as for 200 , we can compute the value for that gives the appropriate : , using a cell doubling time of 55 minutes . With this value for the repressor binding rate , a single repressor molecule in an E . coli cell would take ∼200 s to find a free operator . This is somewhat faster than the 354 s reported by Elf et al . [51] . Using the above value for and the ratio of to from the linear fit of the experimental data we obtained the value for the transcription rate = . This rate for transcription initiation resulted in a steady state concentration of ∼2500 LacY molecules per cell in the fully induced state , within a factor of two of the ∼1000–1200 reported in the literature [22] , [26] . The value also falls within the range of 1000–3000 seen for other highly expressed proteins in E . coli [85] . Accurate measurements of the burst duration in the lac system , particularly in the fully induced state , would increase the accuracy of our model . In order to reproduce a burst frequency of over the mean LacY lifetime in the model , the repressor should dissociate from the operator with a frequency , assuming that each dissociation event produces a burst and that the cell cycle . The burst frequencies inferred by Choi et al . for TMG levels ≤100 are relatively constant with a mean of ∼3 bursts . This corresponds to = . Since the dissociation of a repressor dimer is not thought to be significantly affected by the binding of a single inducer molecule , = . The affinity of a repressor dimer with two bound inducer molecules , however , is thought to be much lower , i . e . , the binding of a second inducer molecule essentially knocks the repressor off of the operator . In the absence of this effect , the response to an increase in inducer concentrations would take a significant fraction of the cell cycle . Elf et al . reported a response time of <60 seconds for addition of IPTG to concentrations from 50 – 1 mM [51] . Therefore , we fit such that the response of the model to increase in IPTG agreed with the published data . The best fit value was obtained for ( shown in Figure 8A ) . The final kinetic rates to be defined were those regarding the binding of TMG to the repressor–operator complex ( Equations 10 & 11 ) . As discussed in Methods , we used the same dissociation rates as for IPTG , leaving only the association rates and , both of which can be derived from the value , which is the inducer concentration at which half of the repressor–operator complexes have a bound inducer . Figure 8B shows the effect of varying on the burst frequency . As approaches , the burst frequency begins to diverge from its expected value . This is due to the increasing occupancy of the O state , which can decay much more quickly into a free operator than the other repressed states; with operator free more often , there are more bursts over the lifetime of a protein . A value of 3 mM for gave the best agreement with the experimental burst frequencies for TMG . Using the derived rates , we performed well-stirred stochastic simulations of the lac model in the absence of LacY positive feedback ( NPF model ) , obtaining the stationary LacY distributions as a function of internal inducer concentration shown in Figure 9 . Compared to the intrinsic noise of the two-state model , the NPF model contains additional noise contributions from the non-constant rates for transitioning between active and inactive transcriptional states . The distributions showed the widest cell-to-cell variability due to the intrinsic noise of the system at intermediate inducer concentrations of 50–400 . At high inducer concentrations the population migrated toward a less variable distribution , as expected . Up to 100 , the population distributions agreed well with those reported by Choi et al . but at 200 the agreement began to break down . This discrepancy at concentrations >100 was caused by two primary factors: the burst duration and the action of inducer knocking repressor off of the operator . Increasing the repressor binding rate would improve the fit by decreasing the duration of each burst , but would cause a large increase in the total number of LacY molecules in the fully induced state , which is not supported experimentally . Alternatively , one could increase the value , causing less inducer instigated dissociation of the repressor–operator complex , but this would decrease the responsiveness of the circuit to addition of inducer , which is also not supported experimentally . Clearly , in order for the model to have greater predictive power , additional features would be necessary . For example , adding a delay between production of mRNA to account for the steps of RNAP open complex formation or more detailed modeling of translation . But lacking the in vivo experimental results to validate any additional complexity , we chose to ignore these effects and analyzed the model as described . The gene regulation function ( GRF ) of an genetic system describes the relation between the activity of a gene and its regulatory control elements [86]–[88] . In the steady state , protein production is balanced by protein degradation/dilution . The mean protein count as a function of the control elements provides a method to analyze a GRF . The mean number of LacY per cell as function of the TMG concentration ( Figure 9D ) and the fraction of time spent in the transcriptionally active state ( Figure 9F ) show the regulatory behavior of the NPF model . We saw a typical sigmoidal regulatory response that was well fit by a Hill equation with an inflection at 312 and a Hill coefficient of 2 . 11 . In a stochastic system , though , the mean rate of gene expression is just one piece of information . As important for a stochastic GRF is how the distribution changes with inducer concentration . The Fano factor ( variance/mean ) provides a measure of the variation of the distribution . For reference , the Fano factor of a Poisson process is 1 . For the NPF model ( Figure 9E ) the Fano factor monotonically increases until 100–200 where it peaks at a value of ∼60 and then begins to decrease ending at a lower value of relative noise than at zero inducer . Next we investigated noise in the inducible genetic switch when the positive feedback regulatory link was active ( PFB model ) . The lacY gene located in the lac operon codes for the integral membrane protein LacY , which actively imports inducer molecules ( lactose/ co-transport ) establishing a positive feedback loop as shown in Figure 1C . The presence of active LacY in the membrane creates a concentration gradient enriching the intracellular environment with inducer molecules relative to extracellular space . For a fixed concentration , the underlying GRF for the lac operon therefore operates not only at an increased inducer concentration but , since the number of LacY is different for each cell , across a distribution of internal inducer concentrations . We calculated the population distributions for the PFB model using well-stirred stochastic simulations at various concentrations . Starting from a stationary population distribution in the absence of inducer , each population of 10 , 000 cells was subject to an instantaneous increase in and simulated for twenty-four hours . Above an concentration of ∼10 , cells in the population began to switch to an induced state in which LacY expression was near its maximum value ( see Figure 10A and B ) . Above ∼25 the transition to full expression was relatively concerted throughout the population . In the range of 10–25 , though , there were two transiently stable subpopulations , one uninduced and the other induced – the overall population was bimodal for a time . To quantify the switching behavior of the population , we classified cells at regular time intervals as uninduced with <400–600 LacY ( best fit for each ) or induced with >1750 LacY . Each subpopulation was then analyzed separately . The mean and variance of the distributions ( Figure 10C ) show that , after an initial response phase , the distribution of the uninduced subpopulation was stable over time . This was true even as the total number of cells in the uninduced population was decreasing as cells within it were switching to the induced state . At intermediate inducer concentrations , the uninduced cell population appeared to reach a stationary distribution from which cells independently and stochastically transitioned to the induced state . In contrast , at higher inducer concentrations the population migrated as a whole in a more downhill-like manner . Noise in a GRF can be expressed in terms of its effect on the phenotypic variance in a population under identical environmental conditions . To compare noise between the NPF and PFB models , we first mapped concentrations to mean concentrations in the uninduced and induced subpopulations ( in the NPF model = ) . We then compared both the mean of the LacY distributions and the Fano factor for the two models . The mean values for the LacY distributions ( Figure 11 ) were similar but the noise in the uninduced subpopulation was significantly higher in the model with positive feedback . Since the underlying GRF is equivalent between the two models , it is the action of the GRF on the distribution of concentrations that gives rise to the increase in intrinsic noise in the PFB model . Having established the well-stirred PFB stationary distribution , we next evaluated the effect of in vivo molecular crowding on the distributions , the PFB+IV model . One obvious reaction subject to spatial effects is the rebinding of the repressor to the operator following an unbinding event . Immediately after unbinding , a repressor is necessarily localized near the operator , i . e . it has a memory of its location . As was shown by van Zon et al . [27] , this memory effect increases the probability of repressor rebinding at very short times compared to a well-stirred approximation . Previous studies only considered the effect of normal diffusion following unbinding but there is an additional effect caused by anomalous diffusion due to in vivo crowding . To investigate repressor rebinding in an in vivo environment , we performed reaction-diffusion simulations of a volume centered on an operator immediately following unbinding of a repressor . We varied the packing density of the approximated in vivo environment to study its effect on rebinding . Figure 12A and B shows that there is an anomalous effect at short time scales ( <1 ms ) . Repressor diffusion at very short time scales is normal at the in vitro rate , but between 1–100 there is a period of anomalous behavior , and at very long time scales repressor diffusion returns to normal diffusion behavior with a lower diffusion coefficient D . Brownian dynamics simulations of proteins in a virtual in vivo environment [39] show a similar anomalous behavior when including only steric constraints with a minimum in the time exponent of ∼0 . 8 for proteins slightly larger than the 75 kDa repressor dimer . When electrostatic effects are included in the Brownian dynamics simulations , however , the apparent diffusion coefficient as well as the anomalous exponent change greatly , so our results should only be considered an upper bound on the in vivo effects . Including further electrostatically driven interactions such as non-specific binding , will increase the anomalous behavior of the repressor . The anomalous behavior of the repressor causes it to spend more time near the operator following unbinding than would be expected for purely Brownian diffusion , leading to more encounters with the operator and a potentially greater probability of rebinding . To measure the change in rebinding probability , we counted the number of repressors that rebound to the operator following unbinding versus the number that escaped into bulk solution , defined here as leaving the simulation volume . As can be seen in Figure 12C , as the density of in vivo crowding increases , the probability of rebinding goes up . Compared to an in vitro unpacked environment at 15% probability of rebinding , at 50% packing the probability of rebinding is ∼24% . The distribution of escape times also broadens ( Figure 12D ) with particles in general taking longer to diffuse away . The anomalous memory effect resulted in the duration of some bursts being significantly shorter than expected . To study the effect of burst duration differences on the stationary LacY distributions in a population , we used our lattice microbe method to generate PFB+IV trajectories of spatially resolved rapid-growth E . coli cells ( see Methods ) . Beginning with the stationary distribution from the well-stirred PFB population , 100 cells were simulated at five internal inducer concentrations for one hour , slightly longer than the duration of a cell cycle ( 55 minutes ) , see Video S1 . Over the course of the simulations , distributions in the in vivo models gradually migrated to lower mean values and lower noise , as can be seen in Figure 13 . Two factors caused this migration: First , the shorter burst durations due to the anomalous diffusion effect described above resulted in fewer proteins being produced per burst and more time spent in the inactive state led to more frequent bursts and less noise . Second , the effective increase in repressor due to the decreased reaction volume . In contrast to spatial effects in an in vitro environment [27] , it appears that in vivo crowding lowers both the mean value and the noise in distributions of observables . Since bacterial cells such as E . coli are known to have packing density changes during different portions of the cell cycle and/or growth conditions , this presents the possibility of measuring these in vivo effects on living cells if the observable distributions can be accurately quantified as a function of the cell cycle or growth conditions . As a first attempt at addressing how changes in the cellular environment due to growth conditions affect gene expression noise , we used CET of E . coli cells under slow growth to build a whole-cell model of an individual bacteria ( Figure 14A ) . Under conditions of slow growth in minimal media E . coli B/r K grows as elongated cylinders with diameter ∼400 nm [89] , which are amenable for CET [50] . The tomograms were used to identify the membrane-enclosed volume of an individual cell along with the three-dimensional position of ribosomes within it . The E . coli B/r K cell under slow growth had only of the volume of typical fast growing cells . A central region of the cell was devoid of ribosomes and inferred to be the location of the condensed nucleoid . We studied the operation of the lac circuit in the slow-growth phenotype ( PFB+IV+CET ) using 100 random replica cells . Each replica used the same experimentally measured cellular geometry and ribosome positions , but a random distribution of other molecules including a condensed chromosome ( see Methods for details ) . Cells were simulated using the lattice microbe method in 15 external TMG , starting with LacY and mRNA counts sampled from the uninduced stationary distribution of the well-stirred PFB model , see Video S2 . Simulations were run for either one hour or until the cell had induced , whichever came first . There were clear differences between the slow- and fast-growth in vivo models . Of the 100 slow-growth cells , 11 induced within one hour whereas only a single fast-growth cell induced in the same time period . Also , the mean number of LacY molecules in the uninduced slow-growth population increased ∼15% over the course of one hour , compared to the fast-growth population which decreased ∼15% . Analysis of the simulation trajectories revealed that the primary cause of the differences in LacY distributions between the slow- and fast-growth models was an increased mean inducer concentration in the smaller cells , 100 versus 42 . For a given number of LacY proteins , the cells with the smaller volume had an increased internal inducer concentration . The increased levels of inducer caused a slight lengthening of the mean duration of free operator events , 68 seconds versus 64 seconds , and a corresponding larger burst size . A bigger change was observed in the mean lifetime of the repressor–operator complex , which decreased to 430 seconds from 730 seconds ( Figure 14B , C ) . The decrease effected an increase in the mean number of transcription bursts per hour , to 4 . 3 from 2 . 6 . The slow-growth model provides a first approximation as to the effect of differences in cellular architecture on stochastic gene expression . The model assumed the same number of repressor molecules for smaller cells , which may not be accurate as repressor is known to regulate its own expression . However , since the largest effect was due to an increased rate of repressor unbinding due to elevated inducer levels , which is independent of repressor concentration , we consider the general result of increased burst frequency and rate of induction in smaller cells to be intriguing . It implies that there might be a difference in the switching properties during the first part of the cell cycle following division when a large burst of LacY would have an increased influence on switching due to the reduced cellular volume . Such an effect could potentially be measured using cell synchronization techniques . Although specific ribosome placement likely also influenced repressor rebinding in the slow-growth model , any differences were overshadowed by the effect of the cell volume change . Nevertheless , in a situation where the placed macromolecules are involved in the reaction kinetics , we anticipate accurate ( non-uniform ) placement will take on much greater importance . Another large difference between the slow- and fast-growth models arose due to the presence of a condensed nucleoid coupled with the smaller cell diameter . In the fast-growth cells the chromosome was assumed to be diffuse and not an obstacle for mRNA diffusion . In the slow-growth cells , the chromosome was randomly placed in the ribosome-excluded region observed in the tomograms and it represented an obstruction for mRNA diffusion . Additionally , the operator was positioned in the center of the fast-growth cells and at the edge of the nucleoid in the slow-growth cells . As can be seen in Figure 14D there was a dramatic increase in localization of mRNA in the slow-growth cells as a result of this arrangement . A recent report of mRNA localization in bacteria [90] suggests that the relative locations of transcription and translation in bacteria may indeed be correlated . If that is generally true , then in systems where the location of protein synthesis affects the reaction kinetics it will be important to know the actual position of the gene in the cell and measurement of the dispersion of the transcripts might be one way to quantify whether the gene is physically located near the site of translation and translocation .
Fitting protein population distributions to gene expression models will be a key step in developing simulations of other stochastic cellular systems with predictive power . Parameters obtained from fitting the distributions will drive the computations . Our stochastic simulations of the inducible lac switch provide an opportunity to test the process of extracting parameters from a population distribution arising from a complex gene expression system using simplified but analytically tractable models . To do so , we fit the stationary population distributions from our simulations to both the burst and two-state models ( Figure 1A & B ) and evaluated their capability to recover the stochastic rate constants used in the simulations ( e . g . , , etc ) . The analysis was performed for each of the different noise variations described above , corresponding to the NPF , PFB , and PFB+IV simulations . We excluded the PFB+IV+CET simulations from this study as they were not performed over a range of inducer conditions . The best fit parameter values were obtained by maximum likelihood estimation using the stationary probability density function ( PDF ) for the burst and two-state models , Equations 12 & 14 in Methods . Fits were performed using 10 , 000 cells for NPF and PFB simulations and 100 cells for PFB+IV simulations . Figure 15A & B show parameter estimates obtained from fitting using the burst model's gamma distribution PDF ( Equation 12 ) . The and parameters ( the B subscript indicates parameters for the burst model ) reliably recover the burst frequency and burst size , respectively , in the NPF simulations at low inducer concentrations , but diverge from the simulation values above ∼100 . This is as expected as the model is only valid when the duration of each burst is short enough that sequential bursts can be considered as occurring independently , <5% of the protein lifetime as shown in Results . In particular the divergence occurs near the switching threshold , making this model most suitable for analyzing the system in the uninduced state with low expression levels . However , the clearness of the biological interpretation for the model parameters as the burst frequency and size make the model extremely valuable over the regime it is valid . Fitting the NPF simulation data to the stationary PDF of the two-state model ( Figure 1B; Equation 14 ) provides good parameter estimates over a wider range of inducer concentrations . The fits are shown in Figure 15C–F for the parameters ( ; the TS subscript indicates two-state ) , ( ) , ( the rate constant for operator activation ) , and ( the rate constant for operator inactivation ) , respectively . As the inducer concentration increases , though , many more cells are required to obtain reliable estimates . Using even 10 , 000 cells , we were unable to obtain good fits for the highest expression levels . At these inducer levels so little time is spent in the inactive state that the difference in likelihood values for different switching rates is insufficient to find a unique maximum using 10 , 000 samples . However , as the time spent in the inactive state approaches zero ( ) the probability distribution approaches a negative binomial distribution without dependence on or , so it is possible to estimate the and parameters in the fully induced state by fitting to a negative binomial . The two-state model therefore appears to be a reasonable method for fitting the NPF simulations . Using the fitting parameters ( along with known or estimated mRNA and protein degradation rates ) , one can readily recover the transcription and translation rates as well as the rates of the operator switching between active and inactive states at a given inducer concentration . Even though switching between active and inactive states in the lac switch is not a first order process – it is controlled by 14 reactions – at a given inducer concentration the steady state switching times are reasonably well-approximated by a single exponential . A further improvement in the two-state model would allow and to depend on the inducer concentration using , e . g . , a Hill function . An analytic solution to such a model would allow extraction of parameters from a multivariate fit using data across all inducer concentrations . However , to the best of our knowledge , the analytic form of such a model has not been derived . Using the steady state distributions from the PFB simulations , neither model achieves good fits . For the two-state model , the and parameters are recovered correctly , but the fits for the and parameters are lower than expected . The poor fit for these parameters is due to noise in the switching rates of the cell population caused by differences in internal inducer concentrations . With positive feedback , it will be very difficult to reliably estimate model parameters from population distributions due to non-linear noise . Fitting to experimental data should be done in the absence of positive feedback , such as by using gene knock-outs to eliminate circuit components responsible for positive feedback . However , if an analytic model were developed including positive feedback effects , comparison of systems with and without these effects could provide estimates of positive feedback parameters , e . g . inducer transport rates . Fits to the PFB+IV simulations as well show deviations from the expected values; in vivo crowding noise changes the parameter fits . For these simulations , an additional source of discrepancies with the models is the non-Poissonian behavior of repressor rebinding – there is a positional memory in the system for a short time following unbinding . In our simulations the effect from in vivo conditions due to excluded volume is modest , but there are other in vivo factors still not accounted for in them , especially non-specific binding as recently reported by McGuffee and Elcock [39] , which would have an even larger effect on repressor rebinding . Also , repressor rebinding most likely occurs via a series of 1D sliding and 3D hopping steps , the effect of which on rebinding in a crowded environment is not known . Accounting for in vivo effects when deriving parameter from experimental population distributions , which would include in vivo noise contributions , will be difficult . Possibly an iterative process of refinement may be required , starting with model estimates and proceeding through multiple rounds of spatial simulation . Overall , it appears that fitting population distributions to the two-state model could prove to be an effective way of obtaining rate constants for stochastic simulations of gene regulation . Single-molecule in vivo fluorescence imaging provides a way to experimentally measure these distributions . Measurements over a range of regulatory conditions could then be used to build a stochastic gene regulation function , provided the actual probability distributions from single-molecule experiments were available at each condition . However , it is important to acknowledge that our simulations did not include a contribution from global extrinsic noise . Noise in our simulations under conditions of high expression approaches Poissonian , as expected from the intrinsic noise of an uncorrelated random process . A recent study has clearly shown , though , that there is a constant level of global extrinsic noise in gene expression in E . coli , maintaining population heterogeneity even at high levels of gene expression [85] . This implies that a way to correct for the global extrinsic noise will be needed in order to fit experimental population distributions at high expression levels . The probability distribution for a stochastic biochemical system to be in a particular state represents the totality of information about the system . From it various measures of the behavior of the system such as the mean first passage time between two states or their relative population at the steady state can be obtained . For models of stochastic gene expression , two relevant reaction coordinates are the number of protein and mRNA molecules in the system . We used our stochastic simulations to reconstruct the two-dimensional probability landscapes ( negative log of the PDF ) of the NPF and PFB models at two external TMG concentrations ( Figure 16 ) . The steady-state landscape of the NPF simulations at 500 inducer shows a bistable mRNA distribution that has been reported by others [25] , [91] . One minima is located near 0 /1700 LacY and the other near 10 /1800 LacY . Note that the stable 10/1800 point does not imply that 180 LacY were produced per , as the degradation rates of the two molecules differ . At 500 inducer , the net time some cells stochastically spend in the inactive state is greater than the typical lifetime of the mRNA bursts . These cells then drift to a zero mRNA abundance . The higher density is caused by an accumulation of these cells near the zero mRNA level until their next mRNA burst pushes them back into a random cycle around the mean mRNA burst size . Interestingly , though , the protein distributions at the two mRNA minimum are different , with the protein abundance being slightly lower in the lower mRNA minimum . This means that the protein and mRNA probability distributions are not completely independent of each other; the joint probability distribution has cross terms . While not a large difference , it is nevertheless possible that the joint protein–mRNA distribution could be used to obtain better parameter fits for the two-state model with fewer cells , if the mRNA counts were known . The bimodal distributions seen in the LacY distributions from the PFB simulations ( Figure 10 ) are recapitulated in the probability landscape for switching . The two-dimensional landscape allows classification of both the uninduced and induced states in terms of their relative protein and mRNA abundances . Additionally , the landscape reveals the transition path for switching from the uninduced to the induced state . One can imagine two possible scenarios for the transition , either the gradual build-up of protein by a series of small bursts , or alternatively , by the random occurrence of a small number of larger bursts . For the lac system with DNA looping , Choi et al . [26] have persuasively argued for the random large bursts as the switching initiator . The switching mechanism of the stochastic system in the absence of looping , however , is not so clear . The probability landscape of our lac model suggests that it is actually the occurrence of several large mRNA bursts , on the order of >10 molecules , in quick succession that is responsible for putting the cell on the path to induction . Cells can spend a significant amount of time in a high LacY but low state without inducing . This behavior is apparent in the cell trajectory plotted in Figure 16B . Switching therefore is a process in which not only a protein threshold must be crossed , but also an mRNA threshold . Our goal with this study was to go beyond previous stochastic simulations of the lac circuit by using information from single molecule protein distributions and experimentally determined cellular architecture to constrain the kinetic parameters and estimate the effect of spatial heterogeneity on the response of the switch . The kinetic model of the inducible lac genetic switch presented in this study illustrates the utility of incorporating single-molecule , single-cell data when modeling cellular biochemical systems . The model was derived using a kinetic framework reproducing a linear relationship between protein burst size and inducer concentration at low concentrations , as has been reported experimentally . Analysis of the linear relationship in terms of inducer–repressor–operator interactions suggests that the stoichiometry of repressor binding is such that repressor dimers with one bound inducer still have significant affinity for the lac operator . Furthermore , single-cell population distributions were used to obtain estimates of the effective rate constants for transcription and repression in the cell . With future increases in performance of the lattice microbe simulation method it should be possible to iteratively refine the kinetic rate constants to account for the effects of cellular architecture , such as we obtained here from CET experiments , and cytoplasmic crowding . Using such in vivo adjusted rate constants the in vivo models should then more accurately reproduce experimental population distributions , which are after all measured under in vivo conditions , than the well-stirred models . The lac model without positive feedback provided a baseline for the noise in the regulation of the lac operon . Intrinsic noise at low gene expression was significantly higher than Poissonian and peaked when the promoter was active 10–30% of the time . The model with positive feedback produced similar mean values for a given intracellular inducer concentration , but the noise was substantially greater . We attribute this effect to the non-linear gene regulatory function operating on a distribution of intracellular inducer levels . Global extrinsic noise in the transcription/translation machinery is a large contributor to population heterogeneity at high levels of expression , but we excluded such noise from the current study . Fitting of data from stochastic simulations of the lac switch with the burst and two-state models of gene expression showed both the potential and limitations of these models to interpret stochastic gene regulation . The burst model described the data well under conditions of low expression , when the gene was active for ≤5% of the mean protein lifetime , but diverged for increasing expression levels . The two-state model better described the data at higher levels of expression , but near full induction the error in the activation and inactivation rates became significant . Additionally , the fits provided estimates of the number of cell measurements necessary to produce reliable parameter estimates . With 50 cells the worst-case relative error was ±90% , but with 200 cells it dropped to ±32% . Fitting to joint mRNA–protein distributions might improve parameter estimation . Fits to data with positive feedback indicated that both models were unable to reliably extract parameters from populations with such feedback . Switching of cells from the uninduced to the induced state was observed in the positive feedback model without DNA looping over a range of low inducer concentrations . During switching , the uninduced population maintained a stable stationary distribution while cells stochastically transitioned to the induced population . The probability landscape showed that both an mRNA and a protein threshold must be crossed for a cell to switch to the induced state . The probability landscape for the DNA looping case is likely different , but additional model states would be required to accurately represent DNA looping . Finally , we have presented what we believe to be the first whole-cell simulations of stochastic gene expression using experimentally obtained cellular architecture . These simulations showed that in vivo conditions can impact the stochastic noise in biological systems . Positional memory of transcription factors following unbinding , amplified by anomalous diffusion due to molecular crowding , introduces non-Poissonian statistics . In the case of our lac model in fast-growth cells , this effect caused a decrease in the mean value of the LacY distribution by up to 10% and its noise by up to 20% , for a given environmental condition . In a slow-growth cell phenotype we saw a large increase in burst frequency due to the smaller cell size , as determined from cryoelectron tomography . From this difference we infer that changes in cellular size and/or shape during the cell cycle can have an impact on stochastic processes . Since spatial noise can vary from cell-to-cell or even during the cell cycle so we consider it a type of extrinsic noise . The necessary computational resources and experimental data are becoming available such that computational biologists should consider adding spatial degrees of freedom into physical models of cellular biochemical networks . | Expressing genes in a bacterial cell is noisy and random . A colony of bacteria grown from a single cell can show remarkable differences in the copy number per cell of a given protein after only a few generations . In this work we use computer simulations to study the variation in how individual cells in a population express a set of genes in response to an environmental signal . The modeled system is the lac genetic switch that Escherichia coli uses to find , collect , and process lactose sugar from the environment . The noise inherent in the genetic circuit controlling the cell's response determines how similar the cells are to each other and we study how the different components of the circuit affect this noise . Furthermore , an estimated 30–50% of the cell volume is taken up by a wide variety of large biomolecules . To study the response of the circuit caused by crowding , we simulate the circuit inside a three-dimensional model of an E . coli cell built using data from cryoelectron tomography reconstructions of a single cell and proteomics data . Correctly including random effects of molecular crowding will be critical to developing fully dynamic models of living cells . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"physics",
"cellular",
"structures",
"stochastic",
"processes",
"mathematics",
"theoretical",
"biology",
"cell",
"biology",
"chemical",
"physics",
"gene",
"expression",
"biology",
"computational",
"biology",
"molecular",
"cell",
"biology",
"biophysics",
"molecular",
"biol... | 2011 | Noise Contributions in an Inducible Genetic Switch: A Whole-Cell Simulation Study |
Sphingolipids , lipids with a common sphingoid base ( also termed long chain base ) backbone , play essential cellular structural and signaling functions . Alterations of sphingolipid levels have been implicated in many diseases , including neurodegenerative disorders . However , it remains largely unclear whether sphingolipid changes in these diseases are pathological events or homeostatic responses . Furthermore , how changes in sphingolipid homeostasis shape the progression of aging and neurodegeneration remains to be clarified . We identified two mouse strains , flincher ( fln ) and toppler ( to ) , with spontaneous recessive mutations that cause cerebellar ataxia and Purkinje cell degeneration . Positional cloning demonstrated that these mutations reside in the Lass1 gene . Lass1 encodes ( dihydro ) ceramide synthase 1 ( CerS1 ) , which is highly expressed in neurons . Both fln and to mutations caused complete loss of CerS1 catalytic activity , which resulted in a reduction in sphingolipid biosynthesis in the brain and dramatic changes in steady-state levels of sphingolipids and sphingoid bases . In addition to Purkinje cell death , deficiency of CerS1 function also induced accumulation of lipofuscin with ubiquitylated proteins in many brain regions . Our results demonstrate clearly that ceramide biosynthesis deficiency can cause neurodegeneration and suggest a novel mechanism of lipofuscin formation , a common phenomenon that occurs during normal aging and in some neurodegenerative diseases .
A hallmark of aging and many neurodegenerative disorders is the neuronal accumulation of storage materials . These deposits include lipofuscin that contain undigested membranes and defective proteins [1] , and/or membrane-free aggregates of misfolded proteins [2] . While the pathological roles of these accrued substances are unclear and may vary between diseases , their sequestration may protect neurons from those components that are otherwise highly toxic in soluble forms [3] . However , evidence also suggests that insoluble storage materials are inherently toxic , and in some circumstances these materials may lead to the inhibition of proteasomal and lysosomal functions , which in turn accelerates their further deposition [4] . In addition to the accumulation of storage materials and impaired protein degradation capacity , changes in cellular homeostasis , including alterations in both simple and complex sphingolipid composition , also occur in the brains of patients with neurodegenerative diseases and in the aging brain [5] , [6] . These highly diverse lipids play important structural and signaling functions in the cell , and mediate cell-cell interaction [7] , [8] . Increases in levels of specific species of ceramide , the simplest sphingolipid , have been found in the brains of Alzheimer's disease ( AD ) patients and a mouse model of AD , and correlate with disease severity [6] , [9] , [10] . Similarly , long-chain and very long-chain ceramide species are increased in the brains of HIV-associated dementia patients [11] . Changes in several sphingolipid classes have been observed in the brains of patients with progressive epilepsy with mental retardation ( EPMR ) , a form of neuronal ceroid lipofuscinosis ( CLN8 ) [12] . Furthermore , sphingolipids have been implicated in Parkinson's disease ( PD ) and Huntington's disease ( HD ) [13] . For example , glucocerebrosidase mutations have been suggested to be risk factors for PD and other Lewy body disorders [14] . Fibroblasts of HD patients and the brain of HD animal models exhibit reduced GM1 ganglioside levels [15] . Lastly , changes in sphingolipids have also been associated with metabolic diseases that are caused by mutations of proteins involved in sphingolipid degradation . Storage of sphingolipids in these diseases results in global impairment of lysosomal function [16] . This in turn blocks lysosomal degradation of defective proteins and organelles . While we are beginning to understand the molecular pathology of sphingolipid-related and other lysosomal storage diseases , the roles of sphingolipid metabolism in the progression of aging and neurodegeneration are still not clear . It was reported that disruption of the Lass2 gene that encodes ceramide synthase 2 ( CerS2 ) , caused myelin degeneration , consistent with the restricted expression of this gene within the brain to oligodendrocytes , and secondary loss of cerebellar granule cells [17] . This finding demonstrates that reduction of sphingolipid levels is indeed pathogenic in the brain , at least in oligodendrocytes . Here we report that deficiency of Lass 1 that encodes CerS1 , a ceramide synthase that is predominantly expressed in neurons [18] , [19] , causes progressive Purkinje cell loss in mice . We also show that CerS1 plays a key role in ceramide biosynthesis in the brain , and loss of this protein dramatically impacts many aspects of sphingolipid homeostasis . Lastly , we find that loss of CerS1 leads to accumulation of lipofuscin that are associated with ubiquitylated proteins in many regions of the brain , suggesting that ceramide biosynthesis is critical for protein and organelle homeostasis . These data demonstrate that lipid biosynthesis defects in neurons can directly cause cell death . Furthermore , our results establish a causal link between lipid biosynthesis deficiencies and lipofuscin accumulation , and may provide a common mechanism for deposition of lipofuscin in aging and neurodegenerative diseases .
The flincher ( fln ) mutation arose spontaneously in a colony of NOD . CB17-Prkdcscid/SzJ mice at The Jackson Laboratory . Mice homozygous for the fln mutation were hyperactive and generally smaller in body size beginning at postnatal day seven . By three weeks of age , the size difference between mutant mice and their littermates was very obvious . Although mutant mice continued to gain weight until reaching maturity , adult mutant mice progressively lost weight , like other ataxic mice . However , the lifespan of mutant mice was comparable to that of their wild type or heterozygous littermates despite their weight loss and severe ataxia . Flincher mutant mice displayed cerebellar ataxia beginning at three weeks , which was concomitant with degeneration of cerebellar Purkinje cells ( Video S1 , Figure 1A and data not shown ) . Neuron loss was progressive , and many Purkinje cells had degenerated in mutant mice by four months of age ( Figure 1B–1D ) . Although the soma of Purkinje cells appeared normal in the mutant mice prior to three weeks of age , at postnatal day twelve ( P12 ) , Purkinje cell dendritic arbors were shorter than those in the wild type cerebellum ( Figure 1E–1F ) . This reduction of dendritic arbor size was more pronounced by P17 ( Figure 1G–1H ) . In addition , the reduced density of calbindin immunostaining in the molecular layer at P17 suggested that the complexity of higher order branches of mutant Purkinje cell dendritic arbors may also be reduced compared to that of wild type dendrites . To identify the molecular defect in fln mutant mice , we crossed fln/fln and BALB/cByJ mice , and performed genome scans on F2 mice using sequence tagged site ( STS ) markers . This analysis localized the fln mutation to Chromosome 8 . Fine mapping using these F2 mice and additional F2 mice from a fln×CAST/EiJ cross narrowed the fln mutation to a 0 . 04 cM ( 0 . 8 Mb ) region between two single nucleotide polymorphisms ( SNPs ) , D8SlacAT1 and rs6348000 ( two recombinants/1984 F2 mice ) , containing 34 known protein-encoding genes ( Figure 2A , Figure S1 and Figure S2A ) . To further define the mutant locus , transgenic mice carrying bacteria artificial chromosomes ( BACs ) containing genes from the fln critical interval were generated and crossed with mutant mice ( Figure S2A ) . Among the three BACs used to make transgenic mice , only BAC RP23-349E13 was able to complement the fln mutant phenotype ( Figure S2B–S2C ) . BAC RP23-349E13 contains nine known genes ( Figure 2B ) and the expression of one of these genes , Lass1 , was greatly reduced in fln/fln brains ( Figure 2C , left panels ) . Sequencing of Lass1 RT-PCR products revealed a single nucleotide deletion in exon 5 , which results in a frameshift mutation that likely leads to nonsense-mediated decay of the Lass1 transcript in fln mutant brains ( Figure 2D , left panels ) . Lass1 is transcribed as a part of an unusual bicistronic transcript that also encodes growth differentiation factor-1 ( GDF1 ) , a member of the TGF-ß family [20] . Thus , as expected , Gdf1 expression is also reduced in fln mutant brains ( Figure 2C , left panels ) . To determine if the fln phenotype is a result of loss of Lass1 or of decreased Gdf1 function , we performed cDNA transgenic complementation experiments . Transgenic mice expressing the Lass1 cDNA under the control of the neuron-specific enolase ( NSE ) promoter were generated and mated with fln mutant mice to generate fln/fln mice carrying the Lass1 transgene . These mice did not develop ataxia nor did they have Purkinje cell degeneration , demonstrating that expression of Lass1 in neurons is sufficient to rescue fln-mediated Purkinje cell degeneration ( Figure 2E–2F and data not shown ) . Therefore , loss of Lass1 , not Gdf1 , function underlies the neuropathology observed in fln mutant mice . Toppler ( to ) , a spontaneous recessive mutation causing Purkinje cell degeneration beginning around three weeks after birth , was also mapped to the middle of Chromosome 8 [21] . Postnatal Purkinje cells in to mutant cerebella also display higher order dendritic branching pattern abnormalities . Based on the striking phenotypic similarity , we proposed that to and fln mutations were allelic and performed complementation tests . Fln/+; to/+ compound heterozygous mice exhibited progressive ataxia beginning at three weeks of age ( data not shown ) . These mice also had Purkinje cell degeneration that was indistinguishable from that was observed in mice homozygous for either the to or the fln mutation ( data not shown ) , indicating that the to mutation also likely disrupts the Lass1 gene . This inference was confirmed by the rescue of the to mutant phenotype by the Lass1 cDNA transgene ( Figure S3A–S3B ) . Although the RNA expression level of the bicistronic Lass1-Gdf1 transcript was not affected by the to mutation ( Figure 2C , right panels ) , sequencing of Lass1 RT-PCR products from toppler mutants revealed a missense point mutation in exon 5 resulting in the change of residue Ala266 to Asp ( Figure 2D , right panel , and Figure S4B ) . These results confirmed that the Lass1 gene underlies the pathological changes in both fln and to mutant mice . The Lass1 gene encodes the ( dihydro ) ceramide synthase CerS1 , one of the six ( dihydro ) ceramide synthases ( CerS1–CerS6 ) in mammals , encoded by Lass1 to Lass6 , respectively [22] . In the de novo pathway of ceramide biosynthesis , these enzymes catalyze the condensation of fatty acyl-CoA and dihydrosphingosine ( dhS , also known as sphinganine ) to produce dihydroceramide ( Figure 3A ) . Dihydroceramide , in turn , is desaturated on the dhS backbone by dihydroceramide desaturase to form ceramide , which is the basic building block for complex sphingolipids . In vitro overexpression studies demonstrated that each of the six mammalian ceramide synthases has different fatty acyl-CoA substrate specificity and saturated C18∶0 fatty acyl-CoA is the preferred substrate for CerS1 [23]–[25] . These studies also demonstrated that CerS1 could also use C16∶0 and C20∶0 fatty acid-CoA , albeit inefficiently , but not the unsaturated C18∶1 fatty acyl-CoA that contains a Δ9 cis double bond [25] . Similar in vitro assays suggested that C18∶0 fatty acyl-CoA can also serve as a substrate for CerS5 , a CerS with broader substrate specificity [25] . However , CerS5 is expressed at a lower level than that of CerS1 in the mouse brain [25] . Together these data suggested that the reduction of CerS1 function in the mouse brain would significantly decrease the activities of ceramide synthases that utilize C18 fatty acyl-CoA . To test this hypothesis , we analyzed ceramide synthase activity with three fatty acyl-CoA substrates , including C18∶0 fatty acyl-CoA , in microsomes prepared from wild type and mutant brain homogenates . The activity for C18 fatty acyl-CoA was reduced drastically in fln/fln brain microsomes , indicating that CerS1 is the major CerS in the brain using C18 fatty acyl-CoA ( Figure 3B ) . No difference in ceramide synthase activity for C16 or C24 fatty acyl-CoA was observed between wild type and fln mutant brain microsomes , confirming the high specificity of CerS1 for C18 fatty acyl-CoA , and suggesting that the activities of other ceramide synthase isoenzymes present in the brain were not affected by the severely reduced CerS1 activity ( Figure 3C and data not shown ) . CerS1 is a multiple transmembrane domain protein and shares a conserved TRAM-LAG1-CLN8 ( TLC ) domain with other CerSs . The fln and to mutations reside between the last two predicted transmembrane domains . This region is within the TLC domain but outside of the LAG1 motif that was previously shown to be indispensible for the catalytic activity ( Figure S4A ) [26] . Mice homozygous for either mutation have very similar phenotypes , suggesting that the Ala266Asp to mutation may also result in a dramatic loss of CerS1 function . To test this possibility , plasmids encoding wild type and mutant CerS1 proteins , each with an N- terminal FLAG epitope , were transfected into COS7 cells . Microsomes prepared from transfected cells were assayed for CerS activity using C18 fatty acyl-CoA as a substrate , and the crude enzymatic activity was normalized to the levels of FLAG-tagged CerS1 . Both fln and to mutations impart almost complete loss of CerS1 catalytic activity , demonstrating that the carboxyl end of the TLC domain , and residue Ala266 in particular , are indispensible for CerS1 function ( Figure 3D ) . To investigate the effect of CerS1-deficiency on sphingolipid homeostasis in the brain , total lipids were extracted from wild type and fln mouse brains , and subjected to liquid chromatography coupled mass spectrometry ( LC-MS ) . The total amount of ceramide was decreased by approximately 50% , confirming that CerS1 is a major CerS in the brain ( Figure 4A ) . In line with the previous report on the preference of CerS1 for C18∶0 fatty acyl-CoA , C18∶0 ceramide was reduced approximately two fold in the mutant samples ( Figure 4B ) . Although previous in vitro data demonstrated that unsaturated C18∶1 fatty acyl-CoA was not utilized by CerS1 [25] , the level of C18∶1 ceramide was decreased more than five fold in the fln mutant brain suggesting that either CerS1 can utilize C18∶1 fatty acyl-CoA in vivo , or the C18∶1 side chain is formed by fatty acyl chain desaturation after ceramide biosynthesis ( Figure 4C ) . Our results also confirm that C18 ceramide species are the dominant ceramide species in the mouse brain . Ceramide is the basic building block of complex sphingolipids , such as sphingomyelin and glycosphingolipids . In agreement with the reduced levels of C18 ceramide species , the amount of complex sphingolipids with C18 fatty-acyl groups was also reduced in the fln/fln brain ( Figure S5A and data not shown ) . Contrary to the reduction of C18 sphingolipids observed in the fln mutant brain , the amounts of other long-chain sphingolipids , i . e . sphingolipids with fatty acyl chain length moiety C14 and C16 , were increased significantly , suggesting a compensatory mechanism ( Figure 4D , Figure S5 and data not shown ) . This compensation was not likely due to an upregulation of CerS5 or CerS6 that utilize C14 and C16 fatty acyl-CoA for ceramide biosynthesis , given that we did not observe an increased ceramide synthase activity with C16 fatty acid CoA in fln/fln brain microsomes ( Figure 3C ) . Rather , increased C14 and C16 sphingolipid levels were most likely the result of excessive substrate availability for CerS5 and CerS6 . The excessive substrate availability for CerS5 may also underlie the similar levels of C18 dihydroceramide observed in wild type and fln/fln brains ( Figure S5C ) . Sphingolipids undergo constant synthesis and degradation . Ceramide can be generated by degradation of complex sphingolipids , and be degraded further by ceramidases to release the sphingoid base , sphingosine . Sphingosine , in turn , can be recycled and used as the sphingoid substrate in the salvage pathway of ceramide biosynthesis , which is also catalyzed by CerSs ( Figure 3A ) . Thus , loss of CerS1 function could result in accumulation of sphingosine , and dhS , the sphingoid substrate of de novo ceramide biosynthesis . Indeed , LC-MS analyses of fln/fln brains revealed a 2–4 fold increase in sphingosine and a greater than 10 fold increase in dhS ( Figure 4E–4F ) . The phosphorylated metabolites of sphingosine and dhS , sphingosine-1-phosphate ( S1P ) and dhS-1-phosphate ( dhS1P ) were also significantly increased in mutant brains ( Figure 4G–4H ) . However , the accumulation of dhS and dhS1P was much more pronounced than that of sphingosine and S1P , suggesting that the de novo ceramide biosynthesis pathway is the dominant pathway for ceramide biosynthesis in the brain . In normal aging brains and under some pathological conditions , undigested storage materials may accumulate in neurons to form autofluorecent deposits , termed lipofuscin . Based on the ultrastructure of lipofuscin , it has been suggested that lipofuscin production is a result of impaired autophagy of organelles , especially mitochondria [27] . Consistent with this hypothesis , mutations of several lysosomal proteins have been linked to the neuronal ceroid lipofuscinoses ( CLNs ) , a group of infantile- to juvenile-onset neurodegenerative diseases that are characterized by abnormal lipofuscin accumulation [28] . However , the recent identification of two CLN proteins , CLN6 and CLN8 , as polytopic membrane proteins on the endoplasmic reticulum suggests that non-lysosomal mechanisms may underlie lipofuscin accumulation in these diseases [29]–[31] . Although CLN8 does not show detectable ceramide synthase activity , ceramide profile changes have been observed in the brains of CLN8 patients [12] . In addition , overexpression of CLN8 or CerSs suppressed the growth phenotype of fibroblasts from patients with CLN9 , the causal gene of which has yet to be identified [32] . These results indicate that defects in ceramide homeostasis may contribute to accumulation of lipofuscin . To directly test whether a deficiency of ceramide biosynthesis can induce lipofuscin accumulation , we examined adult fln and to mutant brains . Although autofluorescent deposits were observed sporadically in neurons of four-month-old wild type mice , deposits were widespread in many brain regions in mutant mice , with the highest abundance in deep cerebellar nuclei , pons , medulla , anterior olfactory nuclei , layer IV to layer VI of the cerebral cortex , and the CA3 region of the hippocampus ( data not shown ) . A similar pattern of deposits was observed in sections of mutant brains that were stained with Luxol fast blue , a compound that stains myelin and lipofuscin ( Figure 5A–5B ) . To ascertain that these observed deposits are indeed lipofuscin , we examined their fine structure . Electron microscopy revealed the presence of many vacuoles and electron dense structures in the cytoplasm of mutant neurons ( Figure 5C ) . Higher magnification revealed membranes and vacuoles within these dense structures , as typically seen in neuronal lipofuscin ( Figure 5D ) . Lipofuscin often contains both lipids and ubiquitylated proteins [1] . To test whether ubiquitylated proteins also accumulate in the lipofuscin in CerS1 mutant brains , we performed immunohistochemistry using an antibody against ubiquitin . Indeed , ubiquitin-positive deposits were observed in the same types of neurons harboring lipofuscin , including cerebral cortical neurons , hippocampal CA3 neurons , and neurons in deep cerebellar nuclei , pons , medulla and anterior olfactory nuclei ( Figure 5E–5H , and data not shown ) . Although the autofluorescent deposits are visible across a wide spectrum of light , the intensity of their fluorescence is lower than that of the fluorescent molecule-conjugated secondary antibodies used for ubiquitin immunofluorescence , allowing simultaneous assessment of lipofuscin and ubiquitin-positive puncta . We found that ubiquitylated proteins and autofluorescent deposits were largely colocalized or closely associated ( Figure 5I–5N ) . While lipofuscin accumulated in many neurons in the mutant brain , we did not observe overt death of neurons other than Purkinje cells , although we cannot rule out more subtle changes in the survival of mutant neurons . However , as evidenced by the hyperactivity of CerS1 mutant mice , it is likely that the function of additional neurons is impaired in these mice . Although lipofuscin deposition was much more pronounced in brains of older mutant mice , we also observed weak autofluorescent deposits containing ubiquitin-positive puncta in the soma and swollen dendrites of a small number of Purkinje cells in 3-week old mutant , but not wild type , mice ( Figure 5O–5Q and data not shown ) . This result suggests that CerS1-deficiency also causes accumulation of lipofuscin-like structures in at least some Purkinje cells .
Alterations of sphingolipids have been observed in many neurodegenerative disorders . However , the contribution of these changes to disease pathogenesis is not clear [5] . We have uncovered two spontaneous mutants with progressive degeneration of cerebellar Purkinje neurons and widespread lipofuscin accumulation . By positional cloning , we identified the mutations as alleles of the Lass1 gene , which encodes ceramide synthase 1 ( CerS1 ) , one of the six ceramide synthases in mammals . Our data demonstrate that deficiency of CerS1/LASS1 leads to dramatic alterations in the levels of sphingolipids in the brain , degeneration of cerebellar Purkinje neurons , and widespread lipofuscin accumulation . CerS1 is specifically expressed in neurons in the brain and has been shown to produce C18∶0 ceramide species in vitro or in cultured cells [18] , [19] , [23]–[25] . In the brain of Lass1 mutant mice , C18 sphingolipid species were significantly reduced , confirming the role of CerS1 in C18 ceramide biosynthesis in vivo . Loss of CerS1 also led to a decrease in total ceramide , in agreement with C18 ceramide as a major ceramide in the adult brain [17] . However these sphingolipid decreases were accompanied by an increase in the steady state levels of C14 and C16 sphingolipid species . In addition , sphingoid bases and their phosphorylated metabolites were drastically increased in the mutant brain . In contrast to previous reports suggesting that ceramide is proapoptotic in vitro and can mediate both stress-induced intrinsic and death receptor-mediated extrinsic apoptosis [8] , [33] , our data suggest that reduction of ceramide or complex sphingolipids results in progressive neuron death , particularly Purkinje cell loss . Intracerebroventricular administration of global ceramide inhibitors has been reported to cause acute neurodegeneration [34] . Contrary to in vitro data , this finding , together with our results , demonstrate that decreases in ceramide synthesis can induce neuron death in vivo . Sphingolipid homeostasis is critically balanced in the cell under normal conditions , and thus either increases or decreases of ceramide could be detrimental to the cell . Alternatively , ceramide species with different fatty acyl chains may play distinct physiological roles . For example , loss of one of the two worm ceramide synthases , which produce different ceramide species , resulted in opposite outcomes in C . elegans under hypoxic conditions [35] . Perhaps C18 ceramide , or some of its derived sphingolipids such as C18-sphingomyelin or C18-cerebrosides , act in a prosurvival fashion in neurons , whereas an increase in other ceramide species , such as C16 ceramide , may be apoptotic [35] . Lastly , Purkinje neuron loss may be due to an increase in the sphingoid bases that normally serve as substrates for CerS1 . Sphingoid bases , particularly their phosphorylated metabolites , are potent signaling molecules and have been proposed to regulate many signal transduction pathways [7] . Thus , elevation of these molecules , particularly dhS and dhS1P , which are increased over ten fold in the CerS1-deficient brain , may impair normal functions of sphingolipid signaling , leading to Purkinje cell death . Consistent with previous reports demonstrating that sphingolipid biosynthesis is crucial for dendritic development in cultured Purkinje or hippocampal neurons [36] , [37] , we observed shortened dendritic arbors and decreased density of calbindin immunostaining in the molecular layer in CerS1-deficient cerebella suggesting that dendritic development of mutant Purkinje cells was abnormal . The role of sphingolipids in dendritic development is not clear , but may involve sphingolipid-rich lipid rafts , which are implicated in activity-dependent dendritic development and maintenance of dendritic spines in in vitro experiments [38] , [39] . Dendritic dysfunction or morphological anomalies are often associated with neurodegeneration [40] . While we cannot rule out a secondary cause for abnormalities in Purkinje cell dendrites in CerS1 mutant mice , it is possible that defects in dendritic development also contribute to death of these neurons . In addition to Purkinje cell death , our data clearly demonstrate that deficiency of ceramide biosynthesis can cause the accumulation of lipofuscin , which is often observed in aging brains and in some neurodegenerative diseases . Lipofuscin is known to contain both lipids and proteins that may originate from membrane bound organelles such as mitochondria [1] . Accumulation of lipofuscin has been linked to reduced lysosomal hydrolytic capacity , based on observations of lipofuscin deposition in mutants with deficiencies in lysosomal proteins [1] , [28] . Both increases and decreases in ceramide have been shown to induce autophagy , suggesting that the balance of sphingolipid homeostasis is important for autophagy and/or lysosomal functions [41] , [42] . Alternatively , conditions that increase the load on lysosomes beyond their capacity may also underlie lipofuscin formation . Given the importance of sphingolipids in protein targeting and as components of membranes , reduced sphingolipid levels may increase the demand on , and/or decrease the capacity of lysosomal function by affecting both lysosomal and non-lysosomal membrane dynamics , and targeting of membrane proteins . Increased load of defective membrane organelles and mislocalized membrane proteins could exceed lysosomal hydrolytic capacity of neurons , resulting in lipofuscin formation . Previous reports suggest ceramide might also be important during aging [43] . One of the yeast genes encoding ceramide synthases , LAG1 ( longevity assurance gene 1 ) , was identified in a screen for genes whose expression changed over yeast replicative cycles [44] . Deletion of LAG1 was associated with an increase of replicative life span in yeast . However , a LASS1 variant exhibiting increased expression , when combined with specific HRAS1 and APOE haplotypes , was suggested to contribute to healthy aging and survival in a human population [45] . This finding suggests that appropriately increased ceramide levels might be beneficial to some cells in aging animals . Inversely , given the specificity of Purkinje cell loss observed in CerS1-deficient mice , decreased ceramide levels appear more detrimental to neurons , and accelerate aging phenotypes , including lipofuscin accumulation . Thus our results demonstrate that in addition to neurodegeneration , alteration of ceramide levels can accelerate some aspects of aging . More research is needed to elucidate the role of ceramide and other sphingolipids in aging and neurodegeneration .
All experiments with mice have been approved by The Jackson Laboratory Animal Care and Use Committee according to relevant national and international guidelines . Oligos used for plasmid construction and genotyping are listed in Table S1 . Genotyping assays are described in detail in Text S1 . The fln and to mutations were maintained on a NOD/SzJ background and a FVB/N background , respectively . Bacterial artificial chromosome ( BAC ) transgenic mouse lines were generated by injection of purified BAC DNA into C57BL/6J pronuclei and maintained on the same background . To generate Lass1 cDNA transgenic lines , Lass1 cDNA was amplified from a mouse cDNA library plasmid pME18-FL3-LASS1 ( Open Biosystems ) with primers LZO455 and LZO456 , and inserted into pNSE-Ex4N1 at the BamHI and SpeI sites , downstream of the neuron-specific enolase 2 gene ( Nse2 ) promoter . The plasmid pNSE-Ex4N1 is a modified version of pNSE-Ex4 , which was described previously [46] , with BamHI , NotI and SpeI sites on an adapter inserted into the HindIII site of pNSE-Ex4 . A SalI fragment containing the Lass1 cDNA , the Nse2 promoter and an SV40 polyadenylation sequence was used for pronuclear microinjection . For expression of full-length wild type LASS1 protein in cultured mammalian cells , the Lass1 coding region was recovered from the pNSE-Ex4N1-LASS1 plasmid by restriction digestion with BamHI and EcoRI , and was inserted into pCMV-3Tag-1A ( Stratagene ) , in frame with a 3xFLAG epitope . For expression of mutant LASS1 proteins , Lass1 fragments were amplified from poly A+ RNA isolated from fln/fln or to/to mutant brains , using primers LZO605 and LZO606 . The FLAG epitope-tagged wild type LASS1 construct was cleaved with BsrGI to swap an internal fragment with the fragments containing the mutations . Homozygous fln mice were crossed with BALB/cByJ mice , and F1 heterozygotes were intercrossed to generate F2 mice . Genome scans using polymorphic microsatellite markers were performed on DNA from ten affected and ten unaffected F2 mice . To fine map the mutation , 951 F2 mice were analyzed using polymorphic microsatellite markers and single nucleotide polymorphisms ( SNPs ) . To further narrow the critical interval , fln/fln and CAST/EiJ were crossed , and F1 heterozygotes were intercrossed to generate F2 mice . 1033 F2 mice were analyzed . Total brain RNA preparation , poly-A+ RNA selection and northern blot analyses were performed as described [47] . To generate northern blot probes , fragments of Lass1 and Gdf1 coding sequences were amplified with oligos LZO463 and LZO426 , and oligos LZO445 and LZO446 , respectively . Immunohistochemical studies were performed as described [48] . Briefly , mice were intracardically perfused and brains were postfixed before dehydration and embedding in paraffin . After antigen retrieval in 0 . 01 M citrate buffer ( pH 6 ) , sections were incubated at 4°C overnight with mouse monoclonal antibody against calbindin D-28 ( Sigma-Aldrich , 1∶1000 ) or rabbit polyclonal antibody against ubiquitin ( DAKO , 1∶200 ) in phosphate-buffered saline with 0 . 03% Triton and 5% normal donkey serum . Colorimetric Calbindin D-28 immunohistochemistry was performed as described [49] . Fluorescent detection was performed with Alexa Fluor-conjugated secondary antibodies ( Invitrogen , 1∶200 ) . Autofluorescence was quenched by incubation with 0 . 01% Sudan Black in 70% ethanol , and some sections were counterstained with Hoescht 33258 . To detect autofluorescent deposits , deparaffinated sections were mounted with Fluoromount-G ( Southern Biotech ) without incubation with Sudan Black . For electron microscopy , mice were intracardically perfused with a mixture of 1 . 2% paraformaldehyde and 0 . 8% glutaraldehyde , and brains were postfixed and processed for transmission electron micrography using standard procedures [50] . COS7 cells were cultured in Dulbecco's Modified Eagle's Medium ( Invitrogen ) supplemented with 10% Fetal Bovine Serum ( Hyclone ) . Transfections were performed using Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's protocol . Medium was replaced with fresh medium 12 hours after the transfection , and cells were harvested 48 hours after transfection . For microsomal preparation from mouse brains , the tissue was homogenized in 20 mM HEPES buffer ( pH 7 . 4 , 2 mM KCL , 2 mM MgCl2 , 250 mM sucrose ) supplemented with protease inhibitors ( Sigma ) with a Tissue Tearor homogenizer ( Biospec Products ) . For microsomal preparation from culture cells , harvested cells were lysed in 20 mM HEPES buffer ( see above ) with a 30-gauge syringe . Tissue homogenates or cell lysates were subjected to centrifugation at 1 , 000×g for 5 min at 4°C to remove unbroken tissue debris or intact cells and nuclei . The supernatants were subjected to centrifugation at 10 , 000×g for 10 min at 4°C to remove mitochondria . The supernatants were ultracentrifuged at 100 , 000×g for 1 h at 4°C to collect microsomes , which were re-suspended in HEPES buffer ( see above ) , and protein concentrations were measured with the Bradford method ( BioRad ) . For in vitro ceramide synthase assays , a 100 µL reaction mixture containing 15 µM C17 sphingosine ( Avanti Polar Lipids ) and 50 µM C16 , C18 or C24 fatty acid CoA ( Avanti Polar Lipids ) in 25 mM potassium phosphate buffer ( pH 7 . 4 ) was pre-warmed at 37°C for 5 min . The enzymatic reaction was initiated by adding microsomes ( 15 µg ) to the reaction mixture , which was incubated at 37°C for 15 min , and the reaction was terminated by adding 2 mL extraction solvent ( ethyl acetate/iso-propanol/water at 60/30/10 v/v/v ) , supplemented with d13/C16 ceramide and d13/C22 ceramide as internal standards for mass spectrometry analyses . Lipid extractions and LC/MS analyses were performed as described previously [51] . Briefly , samples were fortified with internal standards . Lipids were extracted twice with 2 ml ethyl acetate/isopropanol/water ( 60/30/10 v/v ) solvent system , dried under a stream of nitrogen , re-suspended into 150 µL 1 mM NH4COOH in 0 . 2% HCOOH in methanol , and analyzed by LC/MS . LC/MS analyses of sphingolipids were performed on a Thermo Finnigan TSQ 7000 triple quadrupole mass spectrometer , operating in a Multiple Reaction Monitoring positive ionization mode . Sphingolipid levels in the brain homogenates were normalized to protein concentration measured by the Bradford method . Protein samples were separated on SDS gels ( BioRad ) and transferred to a nitrocellulose membrane ( Amersham Biosciences ) using standard techniques [52] . FLAG-tagged proteins were detected using M2 mouse monoclonal antibody against FLAG ( Sigma ) . Mouse monoclonal N+/K+-ATPase antibody ( Abcam ) was used as a loading control . Primary antibodies were detected with an appropriate secondary antibodies conjugated with horseradish peroxidase , and detected with ECL ( Amersham Pharmacia ) . In vitro ceramide synthase assays and mass spectrometry data were analyzed with Sigma Plot . Paired t-tests were performed . | Lipids play many essential cellular roles as structural components of biological membranes or signaling molecules . Alterations of lipids have been observed in the brains of patients with neurodegenerative diseases . However , whether these changes can cause neurodegeneration or otherwise influence the pathology of these diseases is unclear . We identified mouse mutations in a gene encoding a neuronally expressed enzyme that generates ceramide , the basic structural component of many lipids known as sphingolipids . These mutations result in progressive ataxia and loss of cerebellar Purkinje cells . In addition , many neurons in these mutant mice harbor lipofuscin , a storage material containing both membranes and proteins , that is present in aging brains and in brains of patients with neurodegenerative disorders , suggesting that both membrane and protein homeostasis are impaired in mutant neurons . This study directly demonstrates that disruption of sphingolipid biosynthesis can lead to selective neuron death and the abnormal accumulation of lipofuscin , and it underscores the need for further study of the roles of lipids in neurodegenerative disorders . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"neurobiology",
"of",
"disease",
"and",
"regeneration",
"lipids",
"animal",
"genetics",
"sphingolipids",
"genetics",
"biology",
"neuroscience",
"genetics",
"and",
"genomics"
] | 2011 | A Deficiency of Ceramide Biosynthesis Causes Cerebellar Purkinje Cell Neurodegeneration and Lipofuscin Accumulation |
Low- and middle-income countries are facing a dual disease burden with infectious diseases ( e . g . , gastrointestinal tract infections ) and non-communicable diseases ( e . g . , diabetes ) being common . For instance , chronic parasite infections lead to altered immune regulatory networks , anemia , malnutrition , and diarrhea with an associated shift in the gut microbiome . These can all be pathways of potential relevance for insulin resistance and diabetes . The aim of this study was to investigate the association between common gastrointestinal tract infections and glycemia in children from non-fee paying schools in South Africa . We conducted a cross-sectional survey among 9- to 14-year-old school children in Port Elizabeth . Stool and urine samples were collected to assess infection status with parasitic worms ( e . g . , Ascaris lumbricoides , Enterobius vermicularis , and Trichuris trichiura ) , intestinal protozoa ( e . g . , Cryptosporidium parvum and Giardia intestinalis ) , and the bacterium Helicobacter pylori . Glycated hemoglobin ( HbA1c ) was measured in finger prick derived capillary blood . All children at schools with a high prevalence of helminth infections and only infected children at the schools with low infection rates were treated with albendazole . The association of anthelmintic treatment with changes in HbA1c 6 months after the drug intervention was also investigated . A high prevalence of 71 . 8% of prediabetes was measured in this group of children , with only 27 . 8% having HbA1c in the normal range . H . pylori was the predominant infectious agent and showed an independent positive association with HbA1c in a multivariable regression analysis ( β = 0 . 040 , 95% confidence interval ( CI ) 0 . 006–0 . 073 , p<0 . 05 ) . No association of HbA1c with either any other infectious agent or albendazole administration was found . The role of H . pylori in diabetes needs confirmation in the context of longitudinal treatment interventions . The specific effect of other gastrointestinal tract infections on glycemia remains unclear . Future studies should integrate the measurement of biomarkers , including immunological parameters , to shed light on the potential mediating mechanisms between parasite infections and diabetes .
In low- and middle-income countries ( LMICs ) , the dual disease burden stemming from infectious diseases ( IDs ) and non-communicable diseases ( NCDs ) poses a challenge to population health and the health systems . Soil-transmitted helminths and schistosomes are estimated to infect over a billion individuals in LMICs [1 , 2] and cause abdominal pain , diarrhea , poor cognitive development , malnutrition , and anemia . As a consequence of such symptoms , school and work performance is affected and physical activity levels compromised [3] . Helminthiases are often chronic , a result of both under-treatment and re-infection . Soil-transmitted helminthiasis , schistosomiasis , and intestinal protozoa infection are intimately connected with poverty , partially explained by lack of clean water , sanitation , and hygiene [4] . NCDs are gaining importance , also in LMICs [5] . For example , the frequency of diabetes mellitus ( DM ) is rising worldwide , and South Africa is among the top five countries in Africa with an estimated DM prevalence of 9 . 2% [6] . This can be attributed primarily to aging , population growth , increasing rates of unhealthy dietary habits , a sedentary lifestyle , and obesity . While NCDs and DM particularly affect older people , it is generally accepted that early life exposures contribute to the accumulation of molecular damage and a higher disease risk later in adulthood [7] . Little is known about how common parasite infections affect glucose homeostasis and DM etiology , particularly at young age . It is conceivable that parasite infections influence the DM risk through different pathways and in opposite directions [8] . On the one hand , parasite-induced alterations of immune regulatory networks , which have evolved to prolong survival in the human intestines , may stimulate anti-inflammatory pathways and decrease the risk of obesity-induced insulin resistance . Malnutrition , diarrhea and , as a result , low body weight related to chronic helminth infections may additionally decrease DM risk . On the other hand , a sedentary lifestyle and anemia have the potential to increase DM risk . Additionally , the mediating role of helminth-induced shifts in the gut microbiome composition remains to be determined [9] . A limited number of recently reviewed epidemiologic studies with inconsistent results investigated the cross-sectional association of different IDs , including lymphatic filariasis [10 , 11] , schistosomiasis [12] , strongyloidiasis [13 , 14] , and soil-transmitted helminthiasis [15 , 16] with DM or insulin sensitivity . In the present study , we followed up on these observations by studying the association of gastrointestinal tract infections due to helminths , intestinal protozoa , and the bacterium Helicobacter pylori with glycated hemoglobin ( HbA1c ) concentration in school children in the frame of the “Disease , Activity and Schoolchildren’s Health” ( DASH ) study in Port Elizabeth , South Africa [3 , 17] . The study provided detailed information on physical activity , fitness , and socioeconomic status ( SES ) to consider as confounding factors on gastrointestinal tract infection status , and intensity of helminth infections to study a possible dose-response relationship; and on the longitudinal course of HbA1c upon selective anthelmintic treatment .
Ethics approval was obtained from ethics committees in both Switzerland ( EKNZ; reference no . 2014–179 , approval date: 17 June 2014 ) , and South Africa ( study number H14-HEA-HMS002 , approval date: 4 July 2014 ) . Written informed consent from the parents/guardians of participating children as well as oral assent from the children were obtained prior to data collection . A total of 1 , 009 grade-4 children aged 9–14 years from eight non-fee paying primary schools were recruited in various parts of Port Elizabeth in the south-eastern part of South Africa , as described before [3 , 17] . The study was part of the 2-year longitudinal DASH study that consisted of three cross-sectional surveys . In each of the cross-sectional surveys , children’s gastrointestinal infections and other health parameters were assessed , including HbA1c , anthropometry , levels of physical fitness , cognitive performance , and psychosocial health . After each survey , helminth-infected children were treated with a single 400 mg oral dose of albendazole . In schools where the prevalence of helminth infection was 20% or above , all children were treated regardless of infection status according to guidelines put forth by the World Health Organization ( WHO ) [18] . Children with other infections ( Cryptosporidium spp . and/or Giardia intestinalis ) in combination with severe symptoms ( e . g . , bloody stool , diarrhea , abdominal pain , and any abnormal lung sounds ) were referred to the nearest local health clinic for individual management . The baseline cross-sectional survey took place in March 2015 . The current study considers data from this baseline survey and the anthelmintic treatment follow-up examination in September/October 2015 . Grade-4 primary school children of the selected schools were included in the study . Children with severe clinical signs and symptoms ( e . g . , severe fever , severe headache , dizziness , nausea , skin rashes , seizures , and diarrhea ) or reported serious health problems , such as Crohn’s disease , liver or kidney diseases , or who participated in any other study were excluded . The SES was derived from housing characteristics and household assets ( S1 Table ) . The SES score of the households was categorized as poorest , poor , and least poor using the scale by Filmer and Pritchett to disaggregate the distribution of the scores [25] . The age of individuals was grouped into five categories ( 8–9 , 10 , 11 , 12 and >12 years ) , according to the age distribution of the population in the study . The body mass index ( BMI ) was calculated as kg/m2 based on the measured height and weight . For physical activity , we used questionnaires on the frequency and duration of certain activities ( how many days in a week the children were physically active for a total of at least 60 min , the traveling time from home to school , and numbers of exercising days and intensity of exercise in children’s leisure time ) . The scores were summed up and equally categorized into tertiles: active , fair , and poor physical activity level according to the distribution of scores . Cardiorespiratory fitness ( VO2 max ) is the maximum rate of oxygen consumption , as measured during incremental exercise . We estimated the individual VO2 max from the 20 m shuttle run test , which is the most widely used field test for determining cardiorespiratory fitness in children [26 , 27] . A complete case analysis was applied . Forty out of 882 participants at baseline moved or changed schools within the 6-month anthelmintic treatment follow-up , and hence , did not participate in the latter cross-sectional survey . Statistical analyses were performed with STATA version 14 . 1 ( StataCorp; College Station , TX , USA ) . Statistical significance was defined as a two-sided p-value<0 . 05 . Descriptive statistics include counts , percentages for categorical variables and , means , and standard deviations ( SD ) for continuous variables . The categorization of DM status by sex is described according to the American Diabetes Association cutoffs for HbA1c . The baseline prevalence of the different gastrointestinal tract infections is presented for the different schools separately . The characteristics of covariates at baseline are presented stratified for infected and non-infected children . To assess the independent association between gastrointestinal tract infections and HbA1c measurement ( treated as continuous numerical data ) at baseline , linear mixed regressions models with random intercepts for schools were computed . Models were a priori adjusted for factors previously shown to be associated with infections and glycemia or diabetes , and therefore with a potential role as confounders: age , sex , SES , Hb , height , weight , BMI , systolic and diastolic blood pressure , physical activity , VO2 max , and body temperature . As a sensitivity analysis , we also omitted weight , BMI , physical activity , VO2 max , anemia and blood pressure from the models as they are potential mediators of infection effects on glycemia or correlates of glycemia . All models were run ( i ) by adding each infection separately without excluding children with other infections; ( ii ) by adding each infection separately and excluding children with other infections; ( iii ) by adding all infection variables simultaneously and; and ( iv ) by adding groups of infections . We also assessed dose-response effects on HbA1c for infectious agents , especially A . lumbricoides , where data on intensity of infection was available . To assess the independent effect of anthelmintic treatment on changes in HbA1c level between baseline and the 6-month anthelmintic treatment follow-up among children from schools without lifestyle intervention and who were infected at baseline , linear mixed regression models with random intercepts for schools were built . Models were a priori adjusted for age , sex , SES , Hb , height , weight , BMI , diastolic and systolic blood pressure , physical activity , VO2 max , and body temperature , considering information from both time points , as appropriate . Longitudinal models were re-run among subjects infected at baseline but not at follow-up , to differentiate between the effect of the anthelmintic treatment itself and the effect of resolved infection on change in HbA1c . Models were also run for children infected with nematodes and for children with any gastrointestinal tract infection separately .
Complete data records including the baseline and 6-month anthelmintic treatment follow-up surveys were available from 842 children ( Fig 1 ) . Fig 2 shows the distribution of HbA1c at baseline and at the 6-month follow-up for the total study sample of 842 children irrespective of the intervention that they obtained . There was a small shift towards lower HbA1c levels at follow-up ( p<0 . 001 ) , reflecting the lifestyle intervention in some schools . The results of quality control tests underline the validity of the HbA1c data . First , HbA1c results did not depend on the day of examination ( p = 0 . 222 ) , body temperature ( p = 0 . 327 ) , or ambient temperature ( p = 0 . 217 ) ( S1 Fig ) . Second , results from the weekly calibration with identical control 1 and control 2 are presented in S2 Table . At baseline , the overall mean HbA1c level of participants was 5 . 79% with SD of 0 . 25 . The prevalence of prediabetes and diabetes according to baseline is presented in S3 Table . A high prevalence of preDM was observed with 605 ( 71 . 8% ) of children having preDM HbA1c levels . Three children ( 0 . 4% ) exhibited HbA1c results ≥6 . 5% at baseline and were offered diagnostic follow-up for DM . The characteristics of the study population and its univariate association with HbA1c are presented in S4 Table . Table 1 shows the prevalence of gastrointestinal tract infections in the study schools at baseline . H . pylori was the predominant infection ( 416 children with a positive RDT result , 49 . 4% ) . At the unit of the school , the prevalence of H . pylori ranged from 27% to 62% . The second most common infections were the soil-transmitted helminths A . lumbricoides and T . trichiura . Two out of eight schools showed very high prevalence of A . lumbricoides infection ( 62 . 5% and 74 . 1% ) , there was a moderate infection prevalence in a third school ( 25 . 9% ) , while the prevalence in the five remaining schools were below 5% . High prevalence of T . trichiura infection was observed in the same two schools where the prevalence of A . lumbricoides prevalence was high ( 66 . 7% and 67 . 9% , respectively ) , while the prevalence of T . trichiura was below 3% in the remaining six schools . In all schools , infection rates were low to very low or even undetectable for intestinal protozoa ( Cryptosporidium spp . 1–5%; G . intestinalis 6–17% ) , the nematode E . vermicularis ( 1–5% ) , and the trematodes S . mansoni ( 1–3%; detected by POC-CCA urine cassette test ) and S . haematobium ( 0% ) . Table 2 simply compares the characteristics of participants with and without a specific gastrointestinal tract infection . Except for H . pylori , the proportion of children with low SES was higher among infected children compared to their non-infected counterparts . Infections with nematodes and G . intestinalis were more common in males , whereas C . parvum infection was more common in females . Infected children were , on average , older than their non-infected peers . Nevertheless , children with an A . lumbricoides , T . trichiura , and E . vermicularis infection had lower height , weight , and BMI compared to non-infected children . However , children infected with A . lumbricoides , T . trichiura , and H . pylori reported higher physical activity , but did not differ with regard to cardiorespiratory fitness . Concerning anemia and HbA1c , no clear pattern of association was evident from the univariate analysis . The results from the multivariable linear regression models of the cross-sectional association of single or grouped infections with HbA1c are presented in Table 3 . We observed a positive association between H . pylori infection and HbA1c , irrespective of adjustments for other infections ( β = 0 . 040; 95% confidence interval ( CI ) 0 . 006–0 . 074 ) . No significant association of HbA1c with any other infectious agent or infection group was observed . Omitting covariates from the multivariable regression models that are potential mediators of infection effects on glycemia ( physical activity , physical fitness , weight , BMI , and anemia ) or correlated outcomes ( blood pressure ) did not materially alter the results presented for the fully adjusted models ( S5 Table ) . Excluding children with DM at baseline or at the 6-month anthelmintic treatment follow-up did not materially alter the findings ( S6 Table ) . In addition , we were not able to show a statistically significant dose-response relationship between intensity of A . lumbricoides and T . trichiura infection and HbA1c levels , albeit adjusted HbA1c levels were highest in children with most intense infections ( S7 Table ) . Results pertaining to the association between albendazole treatment and change in HbA1c level at the 6-month treatment follow-up are presented in Table 4 . The analysis is restricted to children from schools not subjected to lifestyle interventions given the observed slight decrease in HbA1c in the total study sample . Furthermore , only children with any infection or with nematode infection at baseline , respectively , were included . The regression analyses point to statistically non-significant increases in HbA1c concentrations at the 6-month treatment follow-up . The findings from multivariable regression model excluding covariates that could be potential mediators of infection effects on glycemia or correlated outcomes ( weight , BMI , anemia , physical activity , physical fitness , and blood pressure ) point to generally weaker and still statistically non-significant results ( S8 Table ) .
To our knowledge this is the first investigation examining the cross-sectional association of a broad spectrum of gastrointestinal tract infections with glycemia in school-aged children and assessing the impact of anthelmintic treatment on the change in HbA1c values . We observed a positive association between H . pylori infection and HbA1c , while no statistically significant relationship was observed with any other type of infection . Some animal experiments [28] and human epidemiologic studies [10 , 11 , 13 , 15] have shown helminth infections to lower the blood sugar level and inhibit the development of type 1 DM as well as type 2 DM . An inverse relationship between lymphatic filariasis and both type 1 and type 2 DM was reported from India [10 , 11] . Having a previous schistosome infection exhibited a strong protective effect against DM in the People’s Republic of China [12] . Strongyloides stercoralis infection seemed to be associated with a reduced risk of type 2 DM in adult Australians [13] . Soil-transmitted helminth infections were linked with an improvement of insulin sensitivity in Indonesia [15] . Diabetic patients in Turkey were found to have a lower prevalence of parasitic disease than their healthy counterparts [16] . In contrast , a positive association was found between S . stercoralis infection and DM in Brazil , where it was also found that such infections were associated with a high mortality risk among poorly controlled DM patients [14] . A study conducted by Hakim and colleagues reported a high rate of G . intestinalis infection among DM patients [29] . For trematode infections , positive association with HbA1c concentrations were reported from several studies [12 , 30 , 31] . The cross-sectional nature of these studies precludes casual inference . H . pylori is one of the most common human pathogens causing gastrointestinal inflammation . Potential underlying mechanisms linking H . pylori infection and HbA1c levels and DM may include a disturbance of glucose and lipid absorption by the inflamed gastrointestinal tissue . H . pylori infections may also alter host metabolic homeostasis by affecting appetite regulation and energy expenditure through altered balance of ghrelin and leptin secretion , leading to over-eating and metabolic syndrome pathogenesis . The mediating role of gut microbiota alterations remains unknown [32] . The reported associations between H . pylori infection and DM remain inconsistent . The positive association reported among school children in the present study corroborates findings from two large cross-sectional national surveys conducted by Chen and Blaser in American population samples ( one aged ≥18 years and one aged ≥3 years ) and a Taiwanese study in adults , which all found that H . pylori infections were associated with higher mean HbA1c levels [33 , 34] . Several smaller outpatient clinic or hospital based studies in Turkey , Pakistan , and Qatar among adults aged 18 years and above showed a higher prevalence of H . pylori infection in diabetic patients than non-DM control groups [35–37] . Other studies failed to find a positive association between H . pylori and HbA1c or DM [38–40] . In fact , DM patients were found to have higher rates of H . pylori eradication therapy according to national health insurance data from Taiwan . H . pylori eradication treatment success was found to be lower in DM compared to non-DM patients [41 , 42] . Future intervention studies for the treatment of H . pylori should systematically consider changes in glycemia to shed light on the potential etiologic role of H . pylori in DM development . Some studies indicated an improvement of mean HbA1c and insulin resistance in patients with type 2 DM after H . pylori treatment [43 , 44] . We did not observe a statistically significant increase in HbA1c after anthelmintic treatment with albendazole in children harboring nematode infections at baseline , possibly as a result of sample size limitations . The observed direction of the effect is in line with the reported shift towards a Th2 response in helminth-infected individuals . A number of clinical trials with helminth or helminth antigen therapy have reported promising results in inflammatory bowel diseases [45 , 46] , multiple sclerosis [47] , rheumatoid arthritis [48] . After deworming , which triggers several hyper-inflammatory processes and shifts immune responses from Th2 to Th1 , groups of children treated with either albendazole or mebendazole ( against soil-transmitted helminthiasis ) or praziquantel ( against schistosomiasis ) had a higher positive response to the skin-prick test and allergy related symptoms [49 , 50] . Nevertheless , other studies emphasized that anthelmintic treatment did not have an effect on clinical eczema and asthmatic severity scores [51 , 52] . Given that in our study the highest increase in HbA1c after albendazole treatment was observed in children with non-nematode parasite infection , additional research is needed to understand the effect of the anthelmintic drugs on human glucose metabolism . Yet , our results are aligned with the first publication from a randomized placebo-controlled trial in Indonesia , which showed no effect of albendazole treatment on insulin resistance [53] . Our study has several strengths . First , the study population exhibited sufficient prevalence range for at least some of the infectious agents investigated to allow for efficient interrogation of the study objective . Second , the detailed characterization of children allowed us to assess independent associations of parasite infections with glycemia and limiting residual confounding . To analyze the SES of study participants , we chose multiple correspondence analysis ( MCA ) based on household characteristics and assets ownership over more traditional methods thereby minimizing measurement error related to the different calculation methods of income and consumption , recall bias , and seasonal variation of income and expenditure . Third , we used internationally certified HbA1c testing ( Alere Technologies ) , regularly calibrated with standard control procedure . Ehehalt et al . showed that the measurement of HbA1c was a reliable criterion for children and adolescents to diagnose the onset of childhood type 1 DM [54] . In addition , the POC HbA1c test is an accepted screening instrument for pre-DM and type 2 DM [55 , 56] . We carefully evaluated potential measurement error in HbA1c in the light of the observed high prevalence of pre-DM . We demonstrated the absence of correlations with external temperature , body temperature , and examination date . In addition , all models were adjusted for the concentration of Hb , a potentially important confounder , which was assessed with the widely used HemoCue Hb 301 system . We also acknowledge some limitations of our study . Reverse causation remains a problem related to the cross-sectional nature of our main analysis . The low prevalence for some infections limited statistical power for the analyses . The association between H . pylori infection and HbA1c is no longer statistically significant if the p-values are adjusted for the number of infections investigated ( n = 8 ) . Additionally , the co-infections [3] may in part mask opposite effects of different parasites on HbA1c . Examining only one stool sample has a low diagnostic accuracy due to the day-to-day and intra-specimen variation in helminth egg output . To partially remedy this shortcoming , test specificity was increased by preparing duplicate Kato-Katz thick smears from each stool sample . We observed a very high rate of preDM in the children studied , which may limit the generalizability of the observed associations . Despite the fact that the Alere HbA1c testing is minimally affected by hemoglobinopathies , we cannot assess any influence in the absence of genotyping results . Selection bias related to the complete case analysis approach cannot be excluded but the very high participation rate at baseline and the 6-month anthelmintic treatment follow-up ( 5% drop-out rate ) , and the relatively low rate of children not providing stools ( 15% ) are unlikely to have substantially altered the results . In conclusion , the positive cross-sectional association of H . pylori infections with glycemia is consistent with a potential role of this highly prevalent bacterium in DM in LMICs . The direction and causality of the association warrants further scientific inquiry in the context of longitudinal studies and biobanks that focus on specific parasites and integrate immunity as well as other biomarkers to improve mechanistic understanding of parasite-glycemia associations and the potential impact of deworming programs on DM prevalence . | Parasitic worms ( e . g . , pinworm , roundworm , and whipworm ) , intestinal protozoa ( e . g . , Cryptosporidium parvum and Giardia intestinalis ) , and the bacterium Helicobacter pylori persist at high rates in the gastrointestinal tract of people from low- and middle-income countries . These infectious agents are increasingly paralleled by high rates of non-communicable diseases , such as diabetes . We studied the association of glycemia , measured as HbA1c with common gastrointestinal tract infections among school children aged 9–14 years from disadvantaged neighborhoods in Port Elizabeth , South Africa . Our goal was to deepen the understanding of whether specific gastrointestinal tract infections might be early life determinants of elevated HbA1c levels that might lead to diabetes . We found that the bacterium H . pylori was very common among our group of children with a positive association with hyperglycemia . None of the other infectious agents showed such an association . Additional , longitudinal studies are needed to determine whether there is causality for the observed association between H . pylori and hyperglycemia . The integration of biomarkers will allow studying mediating mechanisms . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"education",
"pathogens",
"sociology",
"microbiology",
"social",
"sciences",
"parasitic",
"diseases",
"nematode",
"infections",
"diabetes",
"mellitus",
"hba1c",
"endocrine",
"disorders"... | 2018 | Association between gastrointestinal tract infections and glycated hemoglobin in school children of poor neighborhoods in Port Elizabeth, South Africa |
Chandipura virus ( CHAV ) , a member of the vesiculovirus genus , is an emerging human pathogen . As for other rhabdoviruses , CHAV entry into susceptible cells is mediated by its single envelope glycoprotein G which is both involved in receptor recognition and fusion of viral and cellular membranes . Here , we have characterized the fusion properties of CHAV-G . As for vesicular stomatitis virus ( VSV , the prototype of the genus ) G , fusion is triggered at low pH below 6 . 5 . We have also analyzed the biochemical properties of a soluble form of CHAV-G ectodomain ( CHAV-Gth , generated by thermolysin limited-proteolysis of recombinant VSV particles in which the G gene was replaced by that of CHAV ) . The overall behavior of CHAV-Gth is similar to that previously reported for VSV-Gth . Particularly , CHAV-Gth pre-fusion trimer is not stable in solution and low-pH-induced membrane association of CHAV-Gth is reversible . Furthermore , CHAV-Gth was crystallized in its low pH post-fusion conformation and its structure was determined at 3 . 6Å resolution . An overall comparison of this structure with the previously reported VSV-Gth post-fusion conformation , shows a high structural similarity as expected from the comparison of primary structure . Among the three domains of G , the pleckstrin homology domain ( PHD ) appears to be the most divergent and the largest differences are confined to the secondary structure of the major antigenic site of rhabdoviruses . Finally , local differences indicate that CHAV has evolved alternate structural solutions in hinge regions between PH and fusion domains but also distinct pH sensitive switches . Globally the comparison between the post fusion conformation of CHAV and VSV-G highlights several features essential for the protein’s function . It also reveals the remarkable plasticity of G in terms of local structures .
Chandipura virus ( CHAV; Family Rhabdoviridae , Genus Vesiculovirus ) is an emerging human pathogen associated with deadly encephalitis , principally among children in the tropical areas of India . CHAV can be isolated from natural populations of Phlebotomine sandflies [1–3] and , in the recent years , has caused several outbreaks with high mortality rate [4] . Despite its importance in public health , there are no structural data available on CHAV proteins . In fact , most of the functional and structural information on CHAV has been inferred from studies performed on the prototype vesiculovirus , vesicular stomatitis virus serotype Indiana ( VSVIND ) . This inference is reasonable due to the conservation of the amino acid ( aa ) sequences of the proteins of the two viruses [5]; for instance , aa sequences of glycoproteins G from both viruses share 40% identity and around 65% similarity [6] . The entry of rhabdoviruses into the host cell is mediated by glycoprotein G which constitutes the spikes that protrude at the viral surface [7] . First , G mediates viral attachment via interaction with a cellular receptor that , in the case of VSVIND , has been demonstrated to be the LDL receptor and its family members [8] . Then , after endocytosis [9 , 10] , upon acidification of the inner environment of the endosome , G undergoes a low-pH-induced structural transition from its trimeric pre-fusion form ( PRE ) towards its trimeric post-fusion conformation ( POST ) that drives fusion between the viral and endosomal membranes [11 , 12] . During this conformational change , internal hydrophobic motifs ( the so-called fusion loops ) interact with one or both participating membranes [13] , resulting in their destabilization and merger . Remarkably , low-pH-induced structural transition of rhabdoviral G is reversible [11 , 14] . In fact , there is a pH dependent thermodynamic equilibrium between different states of G . At neutral pH , G exists as a population of mostly monomeric species with a minority of PRE trimers [12 , 15] whereas , at low pH , the G distribution is shifted toward the trimeric POST state [15 , 16] . This is the main difference between rhabdoviral G and other viral fusion glycoproteins activated at low pH for which the PRE conformation is metastable [17 , 18] . Mature glycoprotein G is about 500 aa long ( 495 for VSVIND and 509 for CHAV ) . G has a huge ectodomain ( 452 aa for CHAV-G ) anchored in the lipid bilayer by a single transmembrane domain ( TMD ) which is located upstream a short C-terminal intra-viral segment . The crystal structures of both PRE and POST states of VSVIND-G ectodomain ( VSVIND-Gth , obtained by limited proteolysis of G at the surface of virions of the Indiana Mudd-Summers strain with thermolysin ) have been previously reported [19 , 20] . This revealed that the POST structure of VSV-G has the same fold as the presumptive post-fusion state of glycoprotein gB of herpesviruses [21] and baculovirus gp64 [22] . This defined a new class ( class III ) of fusion proteins [23] . G ectodomain folds into three distinct structural domains [24] ( S1 Fig . ) . A central domain ( CD ) is made of a lateral β-sheet rich region and a central helix which is involved in trimerization of the molecule in both PRE and POST states . A pleckstrin homology domain ( PHD ) is inserted in a loop of CD and the fusion domain ( FD ) is itself inserted in a loop of PHD . The FD is made of an extended β-sheet structure at the tip of which are located the two hydrophobic fusion loops . These rigid domains are connected by segments ( R1–4 ) which , together with the C-terminal segment ( R5 ) connecting CD to TMD , undergo major refolding events during the structural change . During the conformational change from PRE to POST conformation , both FD and the C-terminal part of the ectodomain rotate around CD [20 , 24] ( S1 Fig . ) . As a consequence , both the fusion loops and the TM domain move from one end of the molecule to the other . The FD moves relatively to PHD due to refolding of segments R2 and R3 . There is also a major refolding event in the two segments connecting CD to PHD ( R1 and R4 ) leading to lengthening of the CD central helix F . Similarly , the C-terminal segment of the ectodomain both completely relocates and refolds into α-helix H , which positions itself , in an antiparallel manner , in the grooves of the trimeric core made by the three helices F . This results in the formation of a six-helix bundle that buries acidic amino acids , which constitute pH sensitive molecular switches [19 , 25] . The complex conformational change has been demonstrated to involve monomeric intermediates [15] . In this work , we have characterized the fusion and biochemical properties of CHAV-G and compared them to those of VSV-G . We also determined the structure of a soluble form of CHAV-G ectodomain ( CHAV-Gth , generated by thermolysin limited-proteolysis of recombinant VSV particles in which the G gene was replaced by that of CHAV[26] ) in the POST conformation at 3 . 6 Å resolution . An overall comparison of this structure with VSV-Gth POST , shows a high structural similarity as expected from the comparison of aa sequences . Among the three domains of G , the PHDs appear to be the most divergent and the largest differences are confined to the secondary structure of the major antigenic site of rhabdoviruses . Finally , local differences indicate that CHAV , together with a related group of vesiculoviruses , has evolved alternate structural solutions in hinge regions between PHD and FD and distinct pH sensitive switches .
We compared the fusion activity of both CHAV-G and VSV-G in a cell-cell fusion assay . For this , BSR cells were co-transfected with a plasmid encoding G ( either VSV-G or CHAV-G ) and rabies virus ( RABV ) phosphoprotein fused to the green fluorescent protein ( P-GFP ) . P-GFP is unable to passively diffuse into the nucleus [27] . This allows an easy visualization and counting of multinucleated cells ( syncytia ) formed as a result of the fusion activity of the glycoprotein [25 , 28] . Transfected cells were incubated for 10 minutes with medium adjusted to pH values from 5 to 7 , which was then replaced by medium buffered at pH 7 . 4 . The cells were then further incubated for 1h at 37°C before fixation . We could observe formation of massive syncytia between pH 5 . 0 and 6 . 0 for cells expressing either CHAV-G or VSVIND-G . At pH 6 . 3 , syncytia were smaller for both glycoproteins . At pH 6 . 5 , no syncytia were detected for CHAV-G whereas small syncytia ( less than 20 nuclei ) were still observed for VSVIND-G up to pH 7 . 0 ( Fig . 1 ) . We purified CHAV-G ectodomain from a recombinant VSV in which the G gene was replaced for that of CHAV ( VSV/CHAV-G ) [29] . This chimeric virus is non-pathogenic and can be produced at high titers in BSR cells . At pH 6 . 0 , thermolysin produces a single cut in CHAV-G between residues 419 and 420 ( the cleavage site was determined by mass spectrometry ) . After proteolysis , the CHAV-G ectodomain ( CHAV-Gth , residues 1–419 ) was purified by successive steps of ultracentrifugation and chromatography as previously described [26] . In the case of VSV-G , the pre-fusion trimer , although detected at the viral surface [12] , is not stable in solution and only monomers of G are detected at high pH [11 , 15] . On the other hand , at low pH , the post-fusion trimer is stable [11 , 15] . Monomers and post-fusion trimers of G have distinct sedimentation coefficient and can be easily separated in a sucrose gradient [11 , 25] . Therefore , we analyzed the quaternary structure of CHAV-Gth on linear 5–20% ( w/v ) sucrose gradients at different pHs and compared it with that of VSVIND-Gth . The overall behavior of CHAV-Gth in the sucrose gradients was similar to that of VSVIND-Gth ( Fig . 2A and 2B ) . At pH values from 7 . 0 to 8 . 0 , Gth was found in fractions 9–12 corresponding to monomeric species , while at pH 5 . 5 and 6 . 0 only the trimeric post-fusion form was found in lower fractions of the gradient ( fractions 4–8 ) . However , the pK of the transition ( defined as the pH value at which monomeric and trimeric G are found in equal amounts ) was slightly higher for CHAV-Gth ( pH 6 . 7 ) than for VSV-Gth ( pH 6 . 5 ) . We also investigated the membrane interaction properties of CHAV-Gth by liposome flotation assay . CHAV-Gth/liposome interaction was assessed over a broad pH-range , from pH 5 . 7 to pH 8 . 5 ( Fig . 2C ) . From pH 5 . 7 to 6 . 6 , all the glycoprotein was located in the upper layer indicating a massive association of CHAV-Gth with liposomes . At pH 7 . 0 , the quantity of CHAV-Gth recovered in the bottom fraction increased considerably and only half of the glycoprotein was still detected in association with membranes . From pH 7 . 3 to 8 . 5 , CHAV-Gth was exclusively found in the bottom fraction indicating a poor membrane interacting activity . Finally , when CHAV-Gth-liposome complexes formed at pH 5 . 7 were brought to pH 8 . 5 , membrane flotation experiments revealed a complete dissociation of glycoproteins from liposomes . Taken together , our results indicate that the fusion properties and the associated low-pH-induced conformational change are very similar for CHAV-G and VSV-G . Particularly , as VSV-G [11 , 15] , CHAV-G pre-fusion trimer is not stable in solution and low-pH-induced membrane association of CHAV-G is reversible . We screened several crystallization conditions for CHAV-Gth in presence of n-dodecyl β-maltoside ( DDM ) to avoid rosette formation via aggregation of the protein through its fusion loops . In these experiments , we could obtain CHAV-Gth crystals at pH 4 . 6 and determine the structure of the protein at 3 . 6 Å resolution by molecular replacement . Statistics of data collection and refinement are given in Table 1 . The comparatively low resolution of the structure and starting phases obtained from molecular replacement with a 40% identity model ( only 25% between PH domains ) did not preclude an accurate final model in this case . This is due to a large solvent content and three molecules in the asymmetric unit . This yields over 20 , 000 unique reflections with which to refine three times 3 , 250 atoms per protomer with tight non-crystallographic symmetry and stereochemical restraints and conservative temperature factor models ( see Material and Methods section for more details ) . The structure's accuracy is attested by excellent final statistics , particularly agreement between model and data and stereochemical parameters including the most unbiased indicators ( Table 1 , R-free and Ramachandran plot , respectively ) . Final temperature factors are quite reasonable for this resolution except in the distal parts of fusion domains , in accordance with these parts being the most flexible tip of an elongated molecule ( S2 Fig . ) . The parts omitted in the initial molecular replacement model showed density and could be built de novo after the first few rounds of refinement and rebuilding , and the final composite simulated annealing omit map confirms unambiguous placing of main chain and most side chains except in the distal parts of FD ( S3 Fig . ) . Accordingly , in the following sections , we discuss only those parts of the model , particularly the hydrogen bond networks , for which density is unambiguous . The single trimer in the asymmetric unit has the three protomers in a classic post-fusion conformation . The segments R1 to R5 connecting the different domains were traced in the experimental electron density map . The three chains were built from residues 1 to 415 , 1 to 416 and 1 to 413 , for chains A , B and C , respectively; with internal breaks in segments 26–33 and 27–33 of chain A and C . Those breaks reflect the intrinsic flexibility of segment R1 . The overall structure of CHAV-Gth is very similar to that of VSVIND-Gth in its post-fusion form ( Fig . 3A and B ) [19] . The structural comparison between CHAV-G and VSVIND-G reveals that the three domains are more conserved than would be expected from the sequence divergence [30 , 31] ( Table 2 , Fig . 3C ) . CDs and the upper part of FDs ( excluding the poorly defined extremity of FD ) main chains are almost superimposable within experimental error ( Table 2 , S4 Fig . ) . The PHDs are more divergent than FDs and CDs . Only 79 residues of the PHD ( called PHD core ) out of 90 in CHAV-G are in structurally equivalent positions with VSV-G ( Table 2 ) . The largest differences are confined to helices D and E which are missing in CHAV-G ( see the orange arrow on Fig . 3B ) . A comparison of the sequence of both proteins at these positions reveals that CHAV-G bears an insertion of 4 aa residues at the level of helix E resulting instead in the formation of a small two stranded β-sheet ( Fig . 4 ) . Also , in the equivalent position to helix D , CHAV-G has a deletion of two residues leading to a loop conformation ( Fig . 4 ) . VSV-G segments containing both helix D and helix E are the most exposed part of the glycoprotein in its PRE conformation at the viral surface and , in fact , this region constitutes the major antigenic site of several rhabdovirus glycoproteins [32–35] . This suggests that the structural divergence between CHAV-G and VSVIND-G in this region is the result of the selective pressure from the humoral response of the host immune system . A second distinct feature between CHAV-G and VSVIND-G lies in the hinge R2 and R3 connecting both PHD and FD ( green arrow on Fig . 3B ) . Although the relative orientation of the two domains is the same in both proteins ( Table 2 ) , there are differences in the secondary structure of these hinges . In the POST conformation of VSVIND-G , R2 is mainly a large loop and extends until the first residues of the β-strand c of FD , whereas R3 forms helix C . The contact between the two segments essentially involves hydrophobic residues ( Fig . 5 ) . For CHAV-G , the amino-terminal part of R2 adopts a β-strand conformation , which is followed by a small loop and a small helix; CHAV-G R3 has a structure which is similar to VSV-G R2 , a long loop ending in a β-strand . The interaction between the two segments is mediated by charged and polar residues ( Fig . 5 ) . Despite those differences in R2 and R3 regions , the relative positions of FD and PHD are identical for both glycoproteins ( Table 2 ) . This suggests that the hinge regions need only to be flexible to allow the rotation of the FD . Alignment of G aa sequences for 12 vesiculoviruses ( S5 Fig . ) reveals that the sequences of R2 and R3 segments are rather divergent among vesiculoviruses . This suggests that the structural constraints ( i . e . flexibility and relative position of PHD and FD in the POST conformation ) can be accommodated by very diverse amino acid sequences . Four acidic residues of VSVIND-G ( D268 , D274 , D393 and D395 ) , which are brought close together in the POST six-helix bundle ( Fig . 6A and 6B ) , have been demonstrated to play the role of pH-sensitive switches for the transition back from the POST toward the PRE conformation of VSV-G [25] . The acidic character of the residues corresponding to D274 , D393 and D395 of VSV-G is conserved among the vesiculovirus genus ( S5 Fig . ) . However , this is not the case for the residue corresponding to D268 which is a major pH-sensitive switch of VSVIND-G [25] . Specifically , in CHAV-G , the residue at this position is an alanine ( A271 ) . The absence of this pH-sensitive switch seems to be compensated by CHAV-G’s residues D269 and E234 which form an intra-protomer hydrogen bond with H209 in their vicinity ( Fig . 6D and 6E ) . This clustering is absent for the corresponding , and non-protonable , residues of VSVIND-G POST . Clustering of protonable residues is known to affect the pKa of their side chain . Indeed , when the PropKa program [36] was used to compute the pKa of the side chains of the cluster based on CHAV-G POST structure , it predicted a pKa of about 10 . 5 for the carboxylate of residue D269 . The only other residue with such a pKa shift in CHAV-G POST is residue E399 which corresponds to VSVIND-G D395 which has also been demonstrated to contribute to the pH sensitivity of VSVIND-G POST . Similarly , VSVIND-G D268 was calculated to have a pKa of about 14 . 5 in agreement with its established role as a major pH sensitive switch in VSVIND-G POST . This further suggests the likely role of the CHAV-G POST cluster D269 , E234 and H209 as a pH sensitive switch . The deprotonation of these residues at high pH is certainly destabilizing the POST conformation of CHAV-G . Remarkably , alignment of vesiculoviruses G aa sequences ( Fig . 6 ) reveals that the vesiculoviruses can be classified in two groups . Either they have an aspartate residue in the position corresponding to VSV-G D268 , or they have a histidine in position corresponding to CHAV-G H209 and an acidic residue in position corresponding to CHAV-G E234 and D269 ( Figs . 6C and S5 ) . Unsurprisingly , this segregation which relies only on vesiculovirus glycoprotein local sequences is in agreement with global phylogenetic studies on the genus [37 , 38] . This different switch of CHAV results in a tighter interaction between the base of the six-helix bundle and the PH domain ( Fig . 7 ) . As a consequence , the PHD has not exactly the same orientation relative to CD . This is why the FDs are slightly spread apart in CHAV-G ( Fig . 3B ) . The comparison of VSVIND-G and CHAV-G structures reveal that the domains do not evolve at the same rate and that the major actor on G evolution is the immune system . As a consequence , the most exposed domain in the PRE conformation ( i . e . the PHD ) , which is the target of neutralizing antibodies , is also by far the most divergent domain among vesiculoviruses . This domain , located at the top of the PRE conformation , is also probably involved in receptor recognition . Therefore , it is probable that the humoral immune response is at the origin of receptor diversification among the rhabdovirus family . The principal finding of this work is that , vesiculovirus G have diverged quite far in some respects but retained important functional and structural features . These features can be thus pointed at as essential for the protein’s function . As an example , the relative orientation of CD and PHD is different in CHAV-G and VSV-G POST protomers , but the relative orientation of PHD and FD remains the same despite the sequence and local structure divergence of segments R2 and R3 and the largely different sets of interactions between FD . This highlights the mechanistic importance of maintaining PHD-FD orientation . This also shows that the most important feature for segments R2 and R3 is their flexibility , but not their secondary structure . As a consequence , the structural constraints which exert on these segments are weak and R2 and R3 can tolerate quite different aa sequences . As another example , although pH sensitive switches are absolutely required for the correct function of G , those switches can be located in alternate places , as in the CHAV-G D269-E234-H209 cluster , the counterpart of which does not exist in VSV-G ( Figs . 6C and S5 ) . In conclusion , the vesiculovirus glycoprotein , despite experiencing strong functional constraints , exhibits a remarkable plasticity in terms of sequence and even local structures . This is reminiscent of the remarkable structural conservation of the class II architecture despite the lack of any sequence similarity [39 , 40] . This suggests that evolvability is a key feature for viral fusion glycoproteins: They display a 3D structure that is robust in the face of amino acid sequence variation because of being malleable in local conformations , yet still maintaining the overall trimeric post-fusion conformation that is essential for their function .
CHAV-Gth was purified as previously described [26] . Briefly , a recombinant VSV harboring the CHAV-G gene [29] was produced at high titers in BSR cells , clones of BHK-21 ( baby hamster kidney , ATCC CCL-10 ) . Purified virions were submitted to proteolysis with thermolysin at pH 6 . 0 for 1 . 5h . This treatment generates a soluble fragment of G comprising the residues 1–419 ( Gth ) as assessed by mass spectrometry . After digestion , the virus was pelleted and the viral pellet was suspended in 50 mM HEPES pH 7 . 5 containing 2% glycerol . The suspension containing Gth was clarified by ultracentrifugation at 25000 rpm ( SW35 rotor ) for 1h on a 20% w/v sucrose cushion . Gth purification was achieved on a DEAE trisacryl column ( GE healthcare ) , followed by a size exclusion step using a Superdex S200 10/30 column ( GE healthcare ) . Fractions containing Gth were concentrated up to 10 mg/ml and stored at −80°C until use . BSR cells plated on glass coverslips at 70% confluence were co-transfected with pCAGGS plasmids encoding VSV-G or CHAV-G , and P-GFP plasmid encoding the phosphoprotein of rabies virus fused to GFP [27] . Twenty four hours after transfection , cells were incubated with fusion buffer ( DMEM-10mM MES or HEPES ) at various pHs ( from 5 . 0 to 7 . 0 ) for 10 minutes at 37°C . Cells were then washed once and incubated with DMEM-10mM HEPES-NaOH buffered at pH 7 . 4 , 0 . 1% BSA at 37°C for 1 hour . Cells were fixed with 4% paraformaldehyde in 1× PBS for 15 min . Cells nuclei were stained with DAPI and syncytia formation was analyzed with a Zeiss Axiovert 200 fluorescence microscope at 20x magnification . The oligomeric state of the Gth protein was determined by sucrose gradient centrifugation . 100 μg of VSVIND-Gth or CHAV-Gth were incubated in 200 μl of buffer containing either 50 mM MES-NaOH ( for pH 5 . 5 to 6 . 7 ) or 50 mM Tris-HCl ( for pH 7 to 8 . 5 ) plus 100 mM NaCl at 37°C for 30 minutes . The protein solution was overlaid on a 5-to-20% w/v continuous sucrose gradient in 100 mM NaCl , 50 mM Tris-HCl or 50mM MES-NaOH at the required pHs . Following centrifugation at 35 , 000 rpm for 16 hours in an SW41 rotor , 12 equal fractions were collected manually from the bottom of the gradient . When required , the fractions were dissolved and reconcentrated to the original volume using ultrafiltration devices ( 30 kDa molecular weight cut off ) to get rid of the excess of sucrose . 20 μl of each fraction was analyzed on 10% SDS-PAGE gels stained with Coomassie blue . Membrane flotation assays were performed as previously described [15] . Briefly , 500 μg of fresh liposomes ( phosphatidylcholine: phosphatidylethanolamine: gangliosides , 7:3:1 ) were incubated with 100 μg of purified Gth at the desired pH during 30 minutes at 37°C ( final volume 600 μL ) . The lipid-protein mix was mixed with a 80% w/v sucrose solution so that final concentration of sucrose was around 65% . The mixture was overlaid with 2 mL of 50% w/v sucrose solution and 1 . 5 mL of 10% w/v sucrose solution , both solutions at the desired pH . Cushions were spun for 16 hr at 20°C in a SW55 Beckman rotor at 45 , 000 r . p . m . 20 μL of the top and bottom fraction were analyzed by SDS-PAGE . Gth in the top fraction was considered as liposome-bond , Gth in the bottom fraction was considered as not interacting with the membranes . For the reversibility experiment , lipid-protein mix was re-incubated at pH 8 . 5 after a first stage at pH 5 . 7 . Crystals were obtained at 293 K by the hanging drop vapour diffusion method . Drops were prepared by mixing 1 μl of CHAV-Gth ( 4 mg/ml ) supplemented with 0 . 2% DDM with 1 μl of the reservoir solution ( 12% PEG 3350 , 0 . 1 M sodium acetate pH 4 . 6 ) and equilibrated against 500 μl of reservoir solution . Crystals were harvested in a cryoprotecting solution containing reservoir solution supplemented with 30% glycerol and flash-cooled by plunging in liquid nitrogen . Diffraction data were collected at 100 K at PROXIMA 1 and ID-29 beam lines in SOLEIL and ESRF synchrotrons ( France ) . Data for the best crystal were integrated and scaled using XDS/XSCALE [41] . The crystals belong to space group C2 with cell parameters consistent with 3 molecules in the asymmetric unit and 60% solvent content [42] . The structure was solved by molecular replacement ( MR ) with the PHENIX package [43] . The VSV-G post-fusion protomer ( PDB 2CMZ , chain A ) was first trimmed of non-conserved sidechains using the default procedure in phenix . sculptor . The three search models were the CD , FD and PHD further trimmed of carbohydrates and of a couple of residues upstream and downstream of refolding segments R1 to R5 . The nine expected domains ( corresponding to 3 protomers ) in the asymmetric unit were placed by phenix . phaser . These initial phases were improved by solvent flattening and NCS averaging . Initial maps were clear enough to rebuild CDs and remove those parts of PHDs that did not fit the electron density maps . The model was iteratively rebuiltin COOT [44] and refined with phenix . refine . The refinement protocol was tailored to reduce the parameter to data ratio as much as possible: Tight non-crystallographic symmetry restraints were enforced between the three protomers in the asymmetric unit . B-factors were parameterized initially by TLS domains , allowing subsequent individual B-factor refinement with the very tight restraints on neighboring atoms implemented in recent versions of phenix . refine . This refinement protocol may allow very high B-factors in distal parts of elongated molecules due to the TLS component . The CDs , PHDs and the upper parts of FDs were carefully rebuilt and segments R1 to R5 added in as density became clear enough to model them . No Ramachandran restraints were applied in refinement at any stage to allow orthogonal validation of stereochemistry . All the structure figures were prepared using the program PyMol ( PyMOL Molecular Graphics System . DeLano Scientific LLC , San Carlos , CA , USA . http://www . pymol . org ) . | Chandipura virus ( CHAV , belonging to vesiculovirus genus ) , is an emerging pathogen associated with deadly encephalitis , principally among children , in India . CHAV single envelope glycoprotein G is involved in two successive steps of virus entry: receptor recognition that allows endocytosis of the virion and fusion of viral and cellular membranes that allows release of the viral nucleocapsid into the cytoplasm for the subsequent steps of infection . As for other rhabdovirus , fusion is triggered at low pH and catalyzed by a conformational change of G from its pre-fusion toward its post-fusion conformation . We have determined the crystalline structure of CHAV-G in its post-fusion conformation . An overall comparison of this structure with the previously reported post-fusion conformation of vesicular stomatitis virus G reveals the selective pressure of the humoral immune response: The major antigenic site , located in the most exposed domain of the pre-fusion conformation of the glycoprotein , is also the most structurally divergent region of G . The structure also highlights the remarkable plasticity of G in terms of local structures particularly in some hinge regions and reveals that the vesiculovirus can be classified in two groups based on the nature of their pH sensitive switches which trigger the conformational change of G . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Structure of the Low pH Conformation of Chandipura Virus G Reveals Important Features in the Evolution of the Vesiculovirus Glycoprotein |
Cytomegaloviruses ( CMV ) are highly species-specific due to millennia of co-evolution and adaptation to their host , with no successful experimental cross-species infection in primates reported to date . Accordingly , full genome phylogenetic analysis of multiple new CMV field isolates derived from two closely related nonhuman primate species , Indian-origin rhesus macaques ( RM ) and Mauritian-origin cynomolgus macaques ( MCM ) , revealed distinct and tight lineage clustering according to the species of origin , with MCM CMV isolates mirroring the limited genetic diversity of their primate host that underwent a population bottleneck 400 years ago . Despite the ability of Rhesus CMV ( RhCMV ) laboratory strain 68–1 to replicate efficiently in MCM fibroblasts and potently inhibit antigen presentation to MCM T cells in vitro , RhCMV 68–1 failed to productively infect MCM in vivo , even in the absence of host CD8+ T and NK cells . In contrast , RhCMV clone 68–1 . 2 , genetically repaired to express the homologues of the HCMV anti-apoptosis gene UL36 and epithelial cell tropism genes UL128 and UL130 absent in 68–1 , efficiently infected MCM as evidenced by the induction of transgene-specific T cells and virus shedding . Recombinant variants of RhCMV 68–1 and 68–1 . 2 revealed that expression of either UL36 or UL128 together with UL130 enabled productive MCM infection , indicating that multiple layers of cross-species restriction operate even between closely related hosts . Cumulatively , these results implicate cell tropism and evasion of apoptosis as critical determinants of CMV transmission across primate species barriers , and extend the macaque model of human CMV infection and immunology to MCM , a nonhuman primate species with uniquely simplified host immunogenetics .
Cytomegaloviruses ( CMV ) are large , double-stranded DNA viruses of the family herpesviridae that induce life-long infections . They are found across a broad range of species , and have adapted to their respective hosts over millions of years of co-evolution . For this reason , CMVs are highly species-specific in vivo , with no experimental cross-species transmissions reported in the literature , although anecdotal evidence of accidental cross-species infections exist [1–4] . Some CMV species can replicate to a certain extent in cells of closely related species in vitro [5] , but replication in cells of more distant species is generally prohibited due to the inability of the virus to block apoptosis or innate defense pathways [6 , 7] . CMVs of multiple nonhuman primate species including rhesus macaques ( RM ) and cynomolgus macaques ( CM ) have been described [8 , 9] , opening up invaluable surrogate models for the study of HCMV virulence , pathogenesis , and immunogenicity , as well as CMV-based vaccine development . The rhesus macaque CMV ( RhCMV ) genome has been fully sequenced and annotated [10] . A comparison between the coding potential of RhCMV and HCMV genomes revealed a higher degree of conservation of viral ORFs and gene families between the two species than previously estimated [11] , far exceeding the homology between MCMV and HCMV . In addition , RhCMV has been cloned as a bacterial artificial chromosome ( BAC ) , allowing for genomic manipulation [12] . We have utilized this BAC technology to design RhCMV vectors expressing SIV antigens , and RM vaccinated with these vectors show unprecedented protection against highly virulent SIVmac239 challenge [13 , 14] . This protection is associated with unconventional CD8+ T cell responses that are either MHC-II or MHC-E-restricted [15 , 16] . These unconventional CD8+ T cell responses may be the result of the unique MHC complexity present in RM [17] , or the result of conserved immunoregulatory mechanisms utilized by CMV . In order to parse out the importance of host immunogenetics from strain-specific CMV mechanisms , additional nonhuman primate models of CMV infection are needed . CM are a newly emerging species for the study of CMV . The first two sequences of cynomolgus macaque CMV ( CyCMV ) were recently published , and contain multiple elongations , truncations , and deletions compared to predicted full-length RhCMV genomes [9 , 18 , 19] . Unfortunately , CyCMV has yet to be cloned as a BAC , precluding manipulation of the CyCMV genome for study in CM . Marsh et . al . recently reported that RhCMV 68–1 , the strain used to construct our vaccine vectors , is incapable of infecting Mauritian cynomolgus macaques ( MCM ) , a population with simplified MHC genetics [20–24] . These results cast doubt on CM as a surrogate model for testing of RhCMV vaccine vectors . However , during in vitro propagation RhCMV 68–1 lost multiple genes that may confer additional fitness to RhCMV in the setting of cross-species transmission . First , due to an inversion in the ULb’ region that also resulted in the loss of genetic information on either side of the inverted DNA segment , the surface glycoproteins Rh157 . 5 and Rh157 . 4 ( henceforth referred to by the HCMV homologue names UL128 and UL130 , respectively ) are deleted [25] . This results in the loss of a functional pentameric complex that mediates non-fibroblast cell tropism [26] as well as shedding and horizontal transmission [27] . Second , the same inversion also results in the deletion of three alpha-chemokine-like open reading frames ( ORFs ) of the UL146 family encoded by RhCMV , although the functional consequences of this loss have not been elucidated [25] . Third , a premature stop codon in Rh60/Rh61 ( henceforth referred to by the HCMV homologue name UL36 ) has rendered the anti-apoptotic viral inhibitor of caspase-8 activation ( vICA ) protein nonfunctional [28] . Finally , multiple point mutations have been acquired by the virus during in vitro culture prior to BAC cloning which resulted in premature stop codons and frame shifts in at least three ORFs , Rh13 . 1 ( RL13 ) , Rh152 . /Rh151 ( UL119/UL118 ) and Rh197 ( US14D ) [11] . Importantly , a partially repaired clone of RhCMV 68–1 has been generated , termed RhCMV 68–1 . 2 , in which expression and functionality of the UL36 , UL128 , and UL130 gene products have been restored ( Table 1 , Fig 1 ) [5] . In addition , we have BAC engineered multiple variant RhCMV clones expressing different combinations of UL36 , UL128 , and UL130 , all of which are based on RhCMV 68–1 or 68–1 . 2 ( Fig 1 ) . Here , we set out to determine if repair of these genes would enable cross-species RhCMV infection of MCM . We show that RhCMV 68–1 and RhCMV 68–1 . 2 both replicate in cultured primary MCM fibroblasts , supporting previous reports of cross-species CMV infections in vitro [5] . However , we did not observe RhCMV 68–1 infection of MCM in vivo , supporting the previous findings of Marsh et al [24] . In contrast , RhCMV 68–1 . 2 established a productive infection as measured by generation of transgene-specific T cell immunity and shedding of virus in urine . We further assessed the roles of UL36 , UL128 , and UL130 in the cross-species transmission of RhCMV 68–1 . 2 to MCM , and show that expression of either UL128 together with UL130 , or UL36 alone is sufficient for infection . Thus , we show for the first time that experimental , cross-species RhCMV infection of CM is possible . These results pave the way for utilizing RhCMV vaccine vectors in the CM model , and shed light on the mechanisms restricting cross-species transmission of CMV .
Comparative analyses between non-human primate CMV strains are limited due to a paucity of field isolate sequences published to date . To more thoroughly examine the similarities and differences between RhCMV and CyCMV , we collected urine from Indian-origin RM born and housed at the Oregon National Primate Research Center ( ONPRC ) and MCM that were born and weaned on Mauritius prior to arriving at ONPRC , and isolated CMV from these samples through co-culture on primary rhesus fibroblasts . Co-culturing of urine from three RM ( 19262 , 19936 and 24514 ) resulted in cytopathic effect ( CPE ) consistent with RhCMV infection , and cell lysates of primary rhesus fibroblasts showed strong immunoblotting reactivity with RhCMV specific antibodies ( Fig 2A ) . We employed these same techniques to isolate CyCMV from four MCM , using RhCMV antibodies for immunoblotting of infected primary rhesus fibroblast lysates to detect expression of both immediate-early and late CyCMV proteins ( Fig 2A ) . To acquire full-length CMV isolates with as little in vitro adaptation as possible , passaging of the virus was kept to an absolute minimum , with no more than 4 passages of in vitro culture prior to sequencing analysis . We then deep sequenced purified viral DNA to further evaluate our new field isolates and obtained full genome sequence coverage for all seven novel CMV isolates . To compare our field isolates to other previously described human , nonhuman primate , and rodent CMV strains [8] , we performed phylogenetic analyses using alignments of DNA sequences of the highly conserved DNA polymerase UL54 ( Fig 2B ) , or of the full genome ( Fig 2C ) . While RhCMVs and CyCMVs are closely related to each other , they still clustered separately according to the host species . Our novel RhCMV isolates were grouped with the RhCMV laboratory strains 68–1 and 180 . 92 , whereas all of our MCM CyCMV isolates were closely aligned with the recently published isolate CyCMV Mauritius [19] and demonstrated a close sequence relationship to CyCMV Ottawa , an isolate derived from a Filipino-origin CM [9] . Thus , we have successfully isolated seven novel RhCMV and MCM CyCMV from animals housed at ONPRC . We next examined the coding potential of our RhCMV and CyCMV genomes to identify any adaptations acquired during in vitro passage . Common genomic adaptations have been defined through the study of RhCMV 68–1 , which was passaged extensively in vitro on fibroblasts prior to being cloned as a BAC [11] . These in vitro adaptations were identified by comparing the cloned RhCMV 68–1 BAC sequence with other non-clonal RhCMV sequences , including those acquired directly from the original urine source collected in 1968 [10 , 29] . Interestingly , similar adaptations such as inactivation of UL36 , UL128 or UL130 are frequently observed in HCMV [30–32] . To identify the most probable start and stop codons for all annotated ORFs , we generated a consensus genome sequence across all RhCMV and CyCMV strains and then compared each strain against this consensus sequence . This enabled us to determine if any viral sequence lost genetic information or acquired point mutations leading to frame shifts and premature stop codons in coding regions . A hypothetical full length RhCMV ( FL ) , derived by combining the sequence information of RhCMV 68–1 BAC with primary RhCMV isolate sequences , contains the entire ULb' region and encodes complete versions of the Rh13 . 1 ( RL13 ) , Rh61/Rh60 ( UL36 ) , Rh152/Rh151 ( UL119/UL118 ) and Rh197 ( US14D ) ORFs , all of which are truncated , elongated , or deleted in RhCMV strain 68–1 ( Fig 3 ) . In contrast , RhCMV strain 180 . 92 lacks a genome fragment ranging from Rh159 through Rh166 in addition to frame shift mutations in the Rh06 ( RL11B ) , Rh08 . 1 ( RL11E ) , Rh10 ( vCOX-2 ) , Rh13 . 1 ( RL13 , RL11G ) , Rh21 ( RL11K ) , Rh148 ( UL116 ) , Rh157 . 4 ( UL130 ) , Rh167 ( O14 ) , and Rh220 ( US28F ) ORFs leading to either shortened or elongated coding regions ( Fig 3 ) . Compared to RhCMV 68–1 and 180 . 92 , the new RhCMV isolates closely correspond to the predicted full-length sequence . RhCMV isolate 19262 has frame shifts in seven different genes , whereas RhCMV isolates 19936 and 24514 have frame shifts in three different genes ( Fig 3 , S1 Table ) . We next assessed the coding potential of the new CyCMV isolates , again comparing each strain against an overall CMV consensus sequence . The two published CyCMV isolates ( Ottawa and Mauritius ) have been annotated previously , but unique annotations for each virus make it difficult to compare ORFs across strains and species . Therefore , we re-annotated all CyCMV isolates using the same nomenclature originally introduced for RhCMV [10] , but giving the ORFs a Cy prefix to indicate their CyCMV origin . This approach revealed that both genomes of the two published CyCMV isolates , Ottawa and Mauritius , display a multi-kb loss of genetic information likely due to in vitro adaptation , with CyCMV Mauritius lacking seven ORFs and CyCMV Ottawa lacking six ( Fig 3 ) . Interestingly , both strains have a deletion in a very similar region near the 5’ end of the genome surrounding Cy13 . 1 ( RL13 , RL11G ) . Deletion or mutation of RL13 , a known Fc binding protein [33] , provides HCMV with a growth advantage in vitro [34] . This ORF also appears to be under strong negative selective pressure in macaque CMV , as it is also mutated in RhCMV strains 68–1 , 180 . 92 , and in four of our novel isolates ( Fig 3 ) . This similarity indicates that Rh13 . 1 , Cy13 . 1 , and RL13 may be mechanistically conserved as we have previously proposed [11] . Interestingly , CyCMVs do not encode ORFs that are homologous to Rh08 . 1 ( RL11E ) and Rh22 ( RL11L ) ( Fig 3 ) . The complete genome sequences of primary isolates of RhCMV and CyCMV thus enabled the validation and correction of previous genome annotations . We next calculated the overall inter-strain nucleotide diversity between our CMV isolates . This measure of polymorphism within a population had a value of 0 . 019 ( 19 nucleotide differences per 1000 genomic bases ) across all RhCMV genomic sequences ( Fig 4A ) . This nucleotide diversity is slightly lower than the diversity determined for 100 newly isolated geographically diverse HCMV strains ( 0 . 026 ) [35] , likely due to our limited sample size of RhCMV and the fact that all viruses were isolated from Indian-origin RM housed at ONPRC . Interestingly , the nucleotide diversity across all MCM CyCMV isolates ( including CyCMV Mauritius ) was only 0 . 003 ( Fig 4A ) . It should be noted that the four ONPRC MCMs were imported directly from the island of Mauritius , and should represent active circulating CMV strains from this island . While our sample size of five MCMs lacks the statistical power to completely rule out the presence of additional strains in the entire MCM population , the lower diversity observed in MCM CyCMV compared to RhCMV is in alignment with the natural history of MCM which experienced a severe population bottleneck approximately 400 years ago [36] . Inclusion of CyCMV Ottawa , which was derived from a Filipino-origin CM , increased the level of nucleotide diversity to 0 . 014 ( Fig 4A ) . This nucleotide diversity is similar to comparisons across RhCMV strains , indicating that inter-strain diversity exists in CyCMV , but that it is largely absent in MCM CyCMV isolates . We further assessed the nucleotide diversity across the entire genome to elucidate inter-strain differences ( Fig 4B ) . This analysis revealed that RhCMV diversity is not equally distributed across the genome , but localized to distinct genomic regions . A similar analysis of HCMV indicated that most ORFs of the genome are highly conserved across global isolates while the RL11 gene family , chemokines of the UL146 gene family , and specific viral surface glycoproteins exhibit significant diversity [37] . Similarly , the RL11 gene family in RhCMV shows strain diversity , although less extensive than HCMV , and the viral chemokines of the UL146 gene family are the most diverse ORFs in the genome ( Fig 4B ) . Surprisingly however , the pentameric complex members UL128 , UL130 and UL131a are diverse in RhCMV whereas they show a very high degree of conservation in HCMV ( Fig 4C ) . As discussed above , MCM CyCMV strains show very limited sequence diversity . Moreover , this diversity is highly localized to genes Cy89 , Cy103 and Cy104 , encoding the predicted homologs of HCMV glycoproteins gB , gO , and gH , respectively . CyCMV 31906 encodes for a gB allele that is substantially different from other CyCMVs ( Fig 4D ) , whereas gO and gH in CyCMV 31909 are distinct ( Fig 4E ) . Thus , although the majority of the genomic sequence of MCM CyCMV is highly conserved , the glycoprotein sequences indicate that at least two distinct CyCMV genotypes are present within the MCM population . Similarly , distinct alleles of gB , gO , and gH are used for genotyping of HCMV , with substantially different sequences generally indicating infection by distinct HCMV genotypes [38–40] . Given the high homology in gene content between RhCMV and MCM CyCMV , we tested whether RhCMV was capable of infecting primary MCM fibroblasts . We performed multi-step growth curves on primary MCM and RM fibroblasts infected with RhCMV 68–1 and the partially repaired clone 68–1 . 2 containing intact UL36 , UL128 and UL130 homologs [5] . Both RhCMV clones infected primary MCM fibroblasts and showed similar growth kinetics and peak titers ( Fig 5A ) , indicating that the rhesus virus is able to productively infect cynomolgus cells . Since this finding indicates that RhCMV can overcome cell-intrinsic innate immunity of cynomolgus cells we next examined if RhCMV could also evade adaptive immune recognition in the context of this cross-species infection . CMV encodes multiple immune evasion vectors that powerfully interrupt CD8+ T cell antigen processing and presentation [41] . Indeed , RhCMV and HCMV express four related glycoproteins—Rh182 ( US2 ) , Rh184 ( US3 ) , Rh185 ( US6 ) , and Rh189 ( US11 ) —that act synergistically with very high efficiency to inhibit presentation of MHC class I-restricted epitopes by infected cells [42] . We have previously shown that this MHC class I interference results in the inability of SIVgag-specific CD8+ T cells to recognize fibroblasts infected with RhCMV 68–1 expressing SIVgag [41] . To test whether MHC-I-dependent peptide presentation to CD8+ T cells is similarly altered in RhCMV-infected primary MCM fibroblasts , we measured the ability of an MCM-derived , SIV-specific CD8+ T cell line targeting the Nef103-111RM9 epitope to recognize primary MCM fibroblasts infected with a panel of RhCMV 68–1 clones expressing the SIV Rev-Tat-Nef fusion protein ( RhCMV/rtn ) ( Fig 5B , blue ) . We compared these results to the ability of a RM-derived , SIVnef-specific CD8+ T cell line to recognize primary RM fibroblasts infected with the same panel of RhCMV 68–1 clones ( Fig 5B , red ) . Similar to the previously described inability of SIV-specific T cells to recognize primary RM fibroblasts infected with RhCMV 68–1 expressing SIV antigen [15] , the MCM SIVnef-specific CD8+ T cell line did not respond to primary MCM fibroblasts infected with RhCMV/rtn 68–1 containing an intact Rh182-189 region . In contrast , both the RM and MCM SIVnef-specific CD8+ T cell lines recognized multiple RhCMV/rtn 68–1 mutants that had partial or complete deletions of Rh182-185 . Thus , RhCMV inhibits T cell antigen presentation in both RM and MCM primary fibroblasts . This evasion mechanism of RhCMV is known to promote superinfection [41] , implying that RhCMV should be able to infect CyCMV seropositive MCM . Since RhCMV has a similar gene content to MCM CyCMV , and since RhCMV is capable of infecting and inhibiting CD8+ T cell antigen presentation in primary MCM fibroblasts in vitro we assessed whether RhCMV is capable of infecting MCM by monitoring SIV-specific T cell response longitudinally in MCM inoculated with various RhCMV recombinants expressing SIV antigens as transgenes ( Fig 6A ) . However , upon subcutaneous ( s . c . ) inoculation of four MCM with 1 x 107 PFU of 68–1 RhCMV/gag we did not observe CD4+ or CD8+ T cells responding to overlapping peptide pools covering the SIVgag protein in any of the animals during 6 weeks post-challenge . Since the induction of T cell responses is a highly sensitive indicator of viral infection , this result indicates that RhCMV 68–1 was unable to infect MCM ( Fig 6B , top row ) . This result is consistent with previously published results , which showed lack of immune responses , viral shedding , and viral replication upon inoculation of MCM with RhCMV 68–1 [24] . Given this result , we formed two distinct hypotheses to explain the failure of RhCMV 68–1 to infect MCM . First , all four MCM were previously infected with MCM CyCMV ( Fig 2A ) and we hypothesized that , unlike immunity of RM to RhCMV , immunity against CyCMV could provide protection against RhCMV infection . Second , RhCMV 68–1 has lost the structural genes UL128 and UL130 during passage in vitro , and surface expression of these genes is necessary for formation of the pentameric complex , which mediates viral entry into endothelial and epithelial cells ( Fig 1 ) [43] . In addition , RhCMV 68–1 contains an early truncation in the UL36 ORF that renders the anti-apoptotic properties of this protein nonfunctional . Therefore , we hypothesized that , while lack of these genes still permits superinfection of RM , the limited cellular tropism and reduced anti-apoptosis capabilities of RhCMV 68–1 prevented infection of MCM . To test our hypotheses , we depleted CD8+ cells ( both CD8+ T and NK cells in MCM ) in the same four MCM prior to s . c . challenge with 1 x 107 PFU of 68–1 RhCMV/env and 68–1 . 2 RhCMV/gag ( Fig 6A and 6C ) . We then monitored the CD4+ T cell responses against both SIVenv and SIVgag for 6 weeks post-inoculation . Interestingly , inoculation with RhCMV 68-1/env again failed to induce CD4+ T cell responses despite depletion of CD8+ cells ( Fig 6B; bottom row ) . In contrast , RhCMV/gag 68–1 . 2 induced robust SIVgag-specific CD4+ T cell responses that were maintained through week 6 . This finding ruled out that CyCMV-specific CD8+ T cell immunity protected against infection with RhCMV , and instead indicated that UL36 , UL128 or UL130 , genes missing in RhCMV 68–1 , but repaired in RhCMV 68–1 . 2 , are required for cross-species infection of MCM with RhCMV ( Fig 1 ) . Since RhCMV elicits particularly high frequency T cell responses in the lungs of RM we next assessed SIV-specific CD4+ T cell responses in bronchoalveolar lavages ( BAL ) . We detected SIVgag-specific CD4+ T cell responses in the BAL of all four animals inoculated with 68–1 . 2 RhCMV/gag , but SIVenv-specific CD4+ T cell were below detection limits in MCM inoculated with 68–1 RhCMV/env ( Fig 6D ) . This result shows that T cell immunity at peripheral sites in RhCMV/SIV infected MCM is similar to that found in RM and confirms the absence of RhCMV 68-1-induced immunity . Finally , we assessed shedding of RhCMV in MCM urine ( Fig 6A ) . When urine from various time points was co-cultured on RM fibroblasts followed by immunoblotting of infected cell lysates we detected CyCMV at all time points using anti-IE antibodies . However , SIVenv expression was not observed at any of the time points indicating lack of 68–1 RhCMV/env shedding consistent with the lack of SIVenv-specific T cell responses . In contrast , SIVgag was detected in cell lysates that were co-cultured with urine from all four animals inoculated with 68–1 . 2 RhCMV/gag at 14 weeks post-infection ( Fig 6E ) . These data provide definitive evidence that RhCMV 68–1 . 2 was able to replicate in the MCM host . We next sought to determine whether RhCMV 68–1 . 2 is able to infect fully immunocompetent MCM . To this end , we inoculated a new cohort of four MCM with 68–1 RhCMV/env ( 1 x 107 PFU ) , 68–1 . 2 RhCMV/pol ( 1 x 107 PFU ) , and 68–1 . 2 RhCMV/gag at a reduced dose ( 2 . 5 x 106 PFU ) without prior CD8+ cell depletion . We again observed CD4+ T cell responses to SIVgag and SIVpol expressed by RhCMV 68–1 . 2 , but not to SIVenv expressed by RhCMV 68–1 ( Fig 7A ) . Similarly , all animals developed SIV-specific CD8+ T cell responses against SIV transgenes expressed by 68–1 . 2 ( Fig 7A ) . In accordance with the peripheral blood T cells , we identified large CD4+ T cell responses against SIVgag or SIVpol in the BAL of 68–1 . 2-inoculated animals , but not to SIVenv expressed by 68–1 ( Fig 7B ) . Furthermore , we observed shedding of both RhCMV 68–1 . 2 vectors , but not 68–1 RhCMV/env in the urine from each of the animals ( Fig 7C ) . Taken together , these results demonstrate that RhCMV 68–1 . 2 is capable of replicating in fully immunocompetent MCM . Compared to RhCMV 68–1 , RhCMV 68–1 . 2 contains the genes UL128 and UL130 derived from RhCMV strain 180 . 92 ( Fig 1 , Table 1 ) . UL128 and UL130 encode subunits of the pentameric complex that governs CMV tropism . In addition , UL128 and UL130 show sequence similarities to CCL and CXCL chemokines , respectively , and chemokine activity has been shown for HCMV UL128 in vitro [44] . To determine whether infection of MCM required a functional pentameric complex for RhCMV to overcome species restriction we inoculated four MCM with 68–1 RhCMV/env ( 1 x 107 PFU ) and two of these MCM were co-inoculated with either 68–1 . 2 RhCMVΔUL128/gag ( 2 . 5 x 106 PFU ) or 68–1 . 2 RhCMVΔUL130gag ( 2 . 5 x 106 PFU ) . We observed robust CD4+ and CD8+ T cell responses to SIVgag in the peripheral blood and BAL of all four MCM , regardless of UL128 or UL130 deletion ( Fig 8A and 8B ) . We also observed shedding in the urine of animals inoculated with ΔUL128 and ΔUL130 versions of 68–1 . 2 RhCMV/gag ( Fig 8C ) . In contrast , CD4+ or CD8+ T cell responses against SIVenv were not observed , and 68–1 RhCMV/env was not shed in the urine ( Fig 8A and 8C ) . These results indicated that repair of the pentameric complex in 68–1 . 2 was not the primary factor supporting cross-species infection . In addition to the pentameric complex , a disabling mutation in the gene encoding the anti-apoptotic protein UL36 ( vICA ) was repaired in RhCMV 68–1 . 2 . We therefore also evaluated whether repair of UL36 alone would enable RhCMV 68–1 to infect MCM . We generated RhCMV/gag ( UL36R ) , a recombinant RhCMV 68–1 with a repaired UL36 gene to investigate the role of this protein in RhCMV infection ( Fig 1 ) . In support of our data with RhCMV 68–1 . 2 recombinants expressing intact UL36 , but lacking a pentameric complex , we observed robust CD4+ T cell responses against SIVgag in both the blood and BAL of animals after inoculation with 1 x 107 PFU of 68–1 RhCMV/gag ( UL36R ) ( Fig 8D and 8E ) . Interestingly , although SIVgag-specific CD8+ T cell responses were also observed in the blood , they seemed at a lower frequency ( Fig 8D ) . However , given the small sample size this was not significantly different from 68–1 . 2-inoculated animals . Importantly , we also observed RhCMV/gag 68–1 ( UL36R ) shedding in the urine of all four MCM ( Fig 8F ) . Although RhCMV/gag 68–1 ( UL36R ) lacks the UL128/UL130-containing pentameric complex , RhCMV 68–1 retains the ability to productively infect primary macaque kidney epithelial cells in vitro [45] . Given that kidney epithelial cells are a major reservoir for CMV shedding into the urine , this explains how pentamer deficient RhCMV is found in the urine . Indeed , previous studies have demonstrated that the pentamer-deficient RhCMV strain 68–1 is shed into the urine of rhesus macaques in vivo [41] . Thus , expression of the anti-apoptotic protein UL36 is sufficient to allow RhCMV 68–1 to establish infection in MCM . Infection with 68–1 RhCMVUL36R/gag suggested that UL36 expression was sufficient for cross-species infection of RhCMV 68–1 . To determine whether UL36 was also required for cross-species infection of RhCMV 68–1 . 2 we deleted UL36 from RhCMV 68–1 . 2 generating recombinant 68–1 . 2 RhCMVΔUL36/pol ( Fig 1 ) . In this recombinant we replaced UL36 with SIVpol thus ensuring that any SIVpol-specific T cell responses were generated from UL36-deficient vectors . We inoculated two MCM with 1 x 107 PFU each of 68–1 . 2 RhCMV/gag and 68–1 . 2 RhCMVΔUL36/pol . Surprisingly , we detected CD4+ and CD8+ T cell responses to both SIVgag and SIVpol , indicating that , in the presence of an intact pentameric complex , UL36 is dispensable for infection of MCM by RhCMV ( Fig 8G ) . This conclusion was further supported by the observation that both 68–1 . 2 RhCMV/gag and 68–1 . 2 RhCMVΔUL36/pol were shed in the urine of both animals ( Fig 8H ) . These data indicate that both pentameric complex components and anti-apoptotic proteins independently enable the infection of MCM by RhCMV 68–1 suggesting that species restriction involves multiple layers that each provide a partial , but incomplete barrier to cross-species infection of non-human primate CMVs .
Nonhuman primates are important models of HCMV infection [8] . Here , we have increased our understanding of RhCMV and CyCMV by isolating and sequencing three new RhCMV strains derived from Indian-origin rhesus macaques born and housed at the ONPRC , and four CyCMV isolates derived from cynomolgus macaques originally born on the island of Mauritius . Only two CyCMV strains had been described prior to our present analysis , one from a Filipino-origin and the other from a Mauritius-origin CM [9 , 19] . Unfortunately , neither strain was full-length and both carried deletions of their RL13 homologue and adjoining regions of the genome . In contrast , all four of our MCM CyCMV genomes contain all predicted ORFs , assuming CyCMVs do not express Rh08 . 1 ( RL11E ) and Rh22 ( RL11L ) , an assumption supported by the fact that neither of the previously published CyCMV strains nor any of the four new CyCMV strains published here contain the genetic information for these genes . We also describe the first full-length RhCMV isolate ( 19936 ) that contains all previously predicted ORFs and contains no frame shift mutations outside of the highly variable RL11 gene family [11] . Using our expanded genetic data set , we show that MCM CyCMV field isolates exhibit minimal diversity , in accordance with the sharp genetic bottleneck that occurred in this population approximately 400 years ago [36] . The CyCMV genome has collinear genomic organization and almost identical coding content to RhCMV , but full genome phylogenetic analysis shows distinct and tight clustering of CMV strains between the two species . In addition , the nucleotide diversity across RhCMV and CyCMV genomes is similar if all CyCMV from diverse backgrounds are considered , and also the areas of diversity are similar , indicating that both viral species are closely related . This relationship is mirrored on the host side where the two macaque species are so closely connected , that they naturally interbreed in regions where they co-habitate , producing fertile offspring [46–48] . Thus , our genetic data bridge an important gap in our understanding of similarities and differences between RhCMV and CyCMV strains , and support the use of both RM and CM in future research . Compared to the first complete RhCMV genome ( 19936 ) containing all previously predicted ORFs reported here , RhCMV 68–1 lacks multiple genes and contains multiple point mutations consistent with previous comparisons to a hypothetical complete genome [11] . Despite these defects and in contrast to infection in CM , RhCMV 68–1 is fully capable of establishing and maintaining an infection in RM either naturally infected with RhCMV or experimentally infected with RhCMV 68-1-derived recombinant viruses [49] . However , viremia , shedding , and transmission are clearly reduced for RhCMV 68–1 in RM even during primary infection as compared to low passage isolates of RhCMV , indicating that these mutations are attenuating in vivo [27] . Presently it is not known which mutations need to be repaired to restore the full replication potential of RhCMV 68–1 in RM , but UL128 and UL130 as well as UL36 studied here in the context of MCM infection are likely candidate genes . We demonstrate that lack of CM-infection by RhCMV 68–1 can be overcome by repair of either UL128 together with UL130 , or UL36 . These results suggest that repairing one of the two gene regions renders species restriction due to lack of the other gene region less stringent . A possible explanation for this surprising finding comes from the fact that UL36 and UL128/UL130 perform very different functions in the viral life cycle . UL36 of HCMV and RhCMV , as well as the MCMV homolog M36 , all have been shown to inhibit extrinsic apoptosis by impeding localization of FAS-associated protein with a death domain ( FADD ) to the death-inducing signaling complex ( DISC ) following TNFR1 or CD95/Fas signaling and thus caspase 8 induction [50 , 51] . Indeed , a dominant negative FADD can functionally replace M36 and restore full pathogenicity in vivo [52] . In the absence of UL36 or M36 , CMV-infected cells become more sensitive to FASL or TNF-mediated apoptosis but there is no cell-intrinsic defect [53] . For instance , M36-deficient viruses do not show a growth defect in fibroblasts except in the presence of TNF-producing macrophages [53] . It is conceivable that UL36-deficient RhCMV renders CM cells more sensitive to extrinsic apoptosis signals than RM cells potentially due to non-UL36-mediated , anti-apoptotic mechanisms that can partially compensate for the loss of UL36 in RM , but not in CM . A potential candidate is UL37 ( viral mitochondrial inhibitor of apoptosis; vMIA ) , which binds the pro-apoptotic , mitochondrial-associated protein Bax , preventing mitochondrial permeabilization and release of pro-apoptotic factors in response to intrinsic as well as extrinsic activation [28] . UL36 and UL37 thus act in tandem to suppress apoptosis in CMV-infected cells . Interestingly , UL37 has been previously implicated in controlling species specific cell tropism between human and mouse CMV as insertion of HCMV UL37 enables MCMV to grow in human cells [6] . Moreover , recent evidence indicates that microRNAs expressed by HCMV can exert anti-apoptotic effects [54] , and other mechanisms of CMV apoptosis suppression are currently being explored [55–59] . Thus , RM-specific compensatory mechanism might explain the species-specific dependence on UL36 for infection of CM but not RM . In contrast to UL36 , which acts post-entry , the UL128 and UL130 proteins of HCMV and RhCMV represent subunits of the gH/gL/UL128/UL130/UL131A pentameric complex in the viral envelope that facilitates entry into non-fibroblast cells such as endothelial , epithelial , or myeloid cells [26 , 60] . In both HCMV and RhCMV , repeated passaging in fibroblasts results in selection of pentamer-mutants [25 , 34] . In the case of RhCMV 68–1 , an inversion and loss of genetic information within the ULb' region resulted in deletion of both UL128 and UL130 [25] . RhCMV 68–1 is thus much more limited in the number of different cell types it can infect efficiently in vitro and in vivo [5 , 26] . However , insertion of the UL128 and UL130 genes from RhCMV strain 180 . 92 into RhCMV 68–1 to create RhCMV 68–1 . 2 fully restored efficient infection of endothelial and epithelial cells [5] . It is thus likely that repair of the pentameric complex in RhCMV 68–1 . 2 enabled the virus to broadly enter multiple cell types . Since cell types differ with respect to their susceptibility to apoptosis signals [61] this broadening of cell tropism might thus enable infection of CM even in the absence of a functional UL36 protein . Our finding that both RhCMV 68–1 UL36R and RhCMV 68–1 . 2 can super-infect CyCMV-infected MCM enables the use of this unique nonhuman primate to determine the contribution of host immunogenetics versus strain-specific CMV mechanisms for induction of unconventional MHC-E and MHC-II restricted CD8+ T cells and remarkable protection from SIV replication previously observed in RhCMV/SIV-vaccinated RM [13] . MCM are a particularly attractive group of CM due to a significant population bottleneck 400 years ago , resulting in highly limited immunogenetics [23 , 36 , 62] . Consequently , MCM provide an exceedingly simplified genetic background on which to test various CMV vaccine vectors . Indian rhesus macaques exhibit remarkably complex MHC genetics with each animal expressing up to twenty MHC-I molecules [17] , thereby complicating the study of these SIV-protective , RhCMV-induced CD8+ T cell responses . In contrast , MCM have only seven , completely described MHC haplotypes [23 , 36 , 62] . Therefore , MCM are the ideal nonhuman primate model in which to study the contribution of immunogenetics to RhCMV-induced protective T cell immunity . In addition , our characterization of CyCMV genomes isolated from MCM will facilitate the molecular cloning of CyCMV , thus enabling the comparison of RhCMV and CyCMV-elicited T cell responses within the homologous or heterologous species . Since CM and RM differ with respect to their susceptibility to certain infectious agents , particularly respiratory pathogens like influenza [63] , cross-species comparisons will impact CMV-based vaccine research . Interestingly , the primary MCM isolates exhibited minimal diversity , in accordance with the sharp genetic bottleneck that occurred in this population approximately 400 years ago [36] . Thus , our genetic data bridge an important gap in our understanding of similarities and differences between RhCMV and CyCMV strains , open new avenues of research into understanding the requirements for priming unconventional MHC-E and MHC-II restricted CD8+ T cell responses , and further the utility of both RM and CM in future infectious disease research .
The RhCMV 68–1 BAC has been extensively described [11 , 12] . RhCMV 68–1 . 2 has been previously described and was kindly provided by Dr . Thomas Shenk [5] . The RhCMV 68-1/gag , RhCMV 68-1/env , and RhCMV 68–1 . 2/gag vectors as well as all viruses carrying deletions in the US6 family of T cell evasion genes used in our T cell activation assay have been previously described [15 , 41 , 49] . RhCMV 68–1 . 2/pol was constructed using the same homologous recombination technique previously published for the RhCMV 68-1/pol vector [13 , 64] . Briefly , an expression cassette containing the 5’-region of the SIVmac239 pol protein with an EFIα promotor and a kanamycin ( Kan ) resistance cassette were amplified from a pORI-6K-F5 plasmid ( primer binding site forward: 5’-ACTTAACGGCTGACATG-3’ , primer binding site reverse: 5’-AGCTTAGTACGTTAAAC-3’ ) . The primers used had a 50bp homology to the target region in RhCMV 68–1 . 2 between Rh213 and Rh214 , exactly the same location previously described by us for insertion of other SIV transgenes ( forward primer: 5’-CTGGGTAGTCAACATGGGCATACGAAACTTGCCCGAATAGATGCTCTCAC-3’ , reverse primer: 5’-CTTTTTGGCCAGCGGGTTGGATGATTTCGCGCGTCATGGACTGCTTCACT-3’ ) [49] . Since homologous recombination was carried out in E . coli strain SW105 , the Kan selection marked was removed by heat shock activating the expression of a Flp recombinase . The virus was reconstituted in primary rhesus fibroblasts and transgene expression confirmed by immunoblot . Rh157 . 5 ( UL128 ) and Rh157 . 4 ( UL130 ) deletions were based on the previously described RhCMV 68–1 . 2/gag vector . Primers with 50bp homology arm to regions upstream and downstream of the targeted ORFs ( forward primer UL128: 5’-GTCATGATATAGTTCCGCCTGGCTGTTTAGGCGGCATCCTTCCGGCTAAT-3’ , reverse primer UL128: 5’-ATTTTTCGATAAAAAAATCACAGCAAACATACTGGTTTTACACACTTTAT-3’ ) ( forward primer UL130: 5’-AAAACTATAATCAACAACTCTATACCTTTGTTTTGCTGATGCTATTGCGT-3’ , reverse primer UL130: 5’-ATTAGCCGGAAGGATGCCGCCTAAACAGCCAGGCGGAACTATATCATGAC-3’ ) were designed with primer binding sites to pORI-6K-F5 ( primer binding site forward: 5’-ACTTAACGGCTGACATG-3’ , primer binding site reverse: 5’-AGCTTAGTACGTTAAAC-3’ ) and a Kan resistance cassette flanked by F5-FRT sites was amplified and inserted in place of UL128 or UL130 by homologous recombination . The selection marker was removed by heat shock to activate the expression of a Flp recombinase . Deletion of the targeted ORF was confirmed by PCR and RT-PCR , and the entire viral genome fully sequenced using Illumina deep sequencing as described below . UL36 was repaired in the 68–1 clone RhCMV/gag using galK recombination . Briefly , the galK gene was inserted between nucleotides 831 and 833 ( counted from the ATG start codon ) in the UL36 ORF . In a second recombination step , the galK gene was then replaced with a 100mer oligonucleotide that corresponded to the correct WT sequence ( AACTTGTCAACTAGTACATAGAGTCTGACTAGGAACTCATTTTTTTCTTTACGGAAGCAACCTAGCACCCCGAGCAATTGATTAA ) . Successful recombinations were screened by sequencing of PCR products , and once confirmed , each virus was subsequently sequenced by Illumina deep sequencing as described below . To generate the RhCMV 68–1 . 2/pol ΔRh61/Rh60 ( ΔUL36 ) mutant , the UL36 ORF was replaced by the SIV pol transgene . The same pORI-6K-F5 plasmid containing the SIVmac239 pol protein described above was used as a template , but since the endogenous UL36 promoter was used to drive transgene expression , the PCR product did not contain the EFIα promotor . The primers used to amplify the SIV pol protein had a 50 bp homology to the upstream and downstream region of UL36 ( forward primer 5’-TTGTATATATTGTCGTTATGTGATTTATTGCTACACATCAAATAAACATG-3’ , reverse primer 5’-TCAGTGAACTCAACGTGGTTCGTCAACAAACATAACCTCAGCTTTGTCGT-3’ ) and the following primer binding sites for pORI-6K-F5 ( primer binding site forward: 5’- GTATGTTGTGTGGAATTGTGAG-3’ , primer binding site reverse: 5’-ACCATGCGGGAGGCGTT -3’ ) . After successful recombination , the Kan selection marker was removed as discussed above and transgene expression confirmed by immunoblots of cells infected with the reconstituted virus . To ensure the correct viral genome sequence , viral DNA was analyzed by Illumina deep sequencing as described below . Virus isolation was performed as previously described [65] . Briefly , urine was obtained from late stage SIVmac239-infected rhesus macaques or healthy MCM upon arrival at ONPRC through cystocentesis or following euthanasia . To clarify the urine from cells , debris , and contaminants it was first centrifuged at 2 , 000 x g for 10 minutes at 4°C and then filtered through a 0 . 45 μm filter ( Millipore ) to clear the urine from any bacterial or fungal contamination . Subsequently , primary rhesus fibroblasts were spin-inoculated at 700 x g for 30 minutes at 25°C in 6 well plates by adding 500μl –1 , 000μl of clarified urine per well . Two hours later the urine was washed off the cells with PBS and new DMEM was added . Cells were kept in culture for up to 1 month or until CPE was observed , after which time the cells were harvested and used to infect new primary rhesus fibroblasts to generate a viral stock for further examination . If virus progress was slow , cells were trypsinized and re-seeded to help the infection spread through the entire monolayer . To generate purified viral DNA for next-generation-sequencing , we used a modified Hirt extraction protocol originally designed for the extraction of Polyoma virus DNA from mouse cells [66] . Three T-175 flasks of primary rhesus fibroblasts were infected with our novel CMV isolates at an MOI of 0 . 1 and the viral supernatants were harvested at full CPE after about 7–10 days . Cellular contaminants and residual cells were removed by centrifugation initially at 2 , 000 x g for 10 minutes at 4°C and subsequently at 7 , 500 x g for 15 minutes . The virus was pelleted from the clarified medium by overlaying a sorbitol cushion ( 20% D-sorbitol , 50 mM Tris [pH 7 . 4] , 1 mM MgCl2 ) and centrifuging at 64 , 000 x g for 1 hour at 4°C in a Beckman SW28 rotor . The generated pellet was resuspended in 500μl 10 . 1 TE Buffer ( 10mM Tris , pH 8 . 0; 0 . 1mM EDTA , pH 8 . 0 ) and 500 μl 2x lysis buffer ( 20mM Tris-Cl , pH 8 . 0; 50mM EDTA , pH8 . 0; 200mM NaCl; 1 . 2% w/v SDS ) was added . Finally , to digest the purified virions , 250μg Proteinase K was added and the solution was incubated for 2h at 37°C . The viral DNA was phenol/chloroform extracted twice and precipitated with absolute ethanol at −80°C overnight . The DNA was pelleted for 20 minutes at 13 , 200 x g at 4°C , washed once with 70% ethanol , and subsequently resuspended in water . DNA concentrations were determined using a ND-1000 Spectrophotometer ( NanoDrop Technologies , Inc . ) . Illumina sequencing libraries were generated as previously described [11] . Briefly , DNA was fragmented using an S2 Sonicator and was then converted to libraries using the standard TruSeq protocol . Libraries were examined on a Bioanalyzer ( Agilent ) and the concentration was determined using real time PCR and SYBR Green fluorescence . Next generation sequencing was performed using a MiSeq Next-Generation Sequencing System ( Illumina ) . Libraries were loaded into a MiSeq reagent cartridge at a concentration of 9 pM and single read sequencing was performed for 300 cycles with 6 additional cycles of index reads . The resulting data was imported into Geneious 8 . 1 . 4 . and the sequencing reads were trimmed of all regions exceeding the error probability limit of 0 . 1% to minimize sequencing errors . All reads with a total length of less than 50 bp after quality control were eliminated from further analysis to increase the likelihood of specific alignments during de novo and reference guided assemblies . Genomes of our viral isolates were first de novo assembled using the processed sequencing data , and subsequently all reads were aligned to the generated consensus sequence in a reference guided assembly to examine potential SNPs and ensure accuracy . SNPs that showed a frequency of more than 50% in a location with a depth of at least 10% of the mean depth were considered a true change of the consensus sequence , and hence changes to the consensus sequences were applied and referenced guided assemblies were repeated until no SNP showed a frequency of 50% or more . Full length nucleotide alignments , amino acid alignments and neighbor-joining phylogenetic trees were generated using ClustalW2 [67] . Sliding window analyses and calculations of nucleotide diversity were performed using Variscan , version 2 . 0 . 3 [68] . The sliding window analyses used 100bp windows , incrementing every 50bp . All sequence data were visualized using Geneious , which was also used to generate images of alignments ( Geneious version 9 . 1 , Biomatters Inc . ) . Urine and fibroblast samples were obtained from Indian rhesus macaques ( Macaca mulatta ) housed at the Oregon National Primate Research Center ( ONPRC ) for other ongoing , unrelated studies previously described [15 , 41 , 49] . A total of 12 male and 4 female MCM ( Macaca fascicularis ) originally from Mauritius and housed at the ONPRC were utilized for in vivo infection studies under the approval of Oregon Health and Science University ( OHSU ) Institutional Animal Care and Use Committee ( IACUC ) Protocol 0967 . Primary MCM fibroblasts were a kind gift from Dr . Ole Isacson . All macaques in this study were managed according to the ONPRC animal husbandry program , which aims at providing consistent and excellent care to nonhuman primates . This program is based on the laws , regulations , and guidelines set forth by the United States Department of Agriculture ( e . g . , the Animal Welfare Act and its regulations , and the Animal Care Policy Manual ) , Institute for Laboratory Animal Research ( e . g . , Guide for the Care and Use of Laboratory Animals , 8th edition ) , Public Health Service , National Research Council , Centers for Disease Control , the Weatherall Report titled “The use of nonhuman primates in research” , and the Association for Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) International . The nutritional plan utilized by the ONPRC is based on National Research Council recommendations and supplemented with a variety of fruits , vegetables , and other edible objects as part of the environmental enrichment program established by the Behavioral Management Unit . Paired/grouped animals exhibiting incompatible behaviors were reported to the Behavioral Management staff and managed accordingly . MCM were challenged subcutaneously with varying doses of RhCMV/SIV vectors in the upper arm or upper leg . In cases where a single MCM was challenged with multiple RhCMV/SIV vectors , each vector was administered at a distinct anatomic location ( arms and legs ) . All efforts were made to minimize suffering through the use of minimally invasive procedures , anesthetics , and analgesics when appropriate . Animals were painlessly euthanized with sodium pentobarbital and euthanasia was assured by exsanguination and bilateral pneumothorax , consistent with the recommendations of the American Veterinary Medical Guidelines on Euthanasia ( 2013 ) . Telomerized rhesus fibroblasts were spin-inoculated with centrifuged filter-sterilized ( 0 . 4mm ) urine from both RM and MCM at 1 , 600 x g for 1 hour at 4°C . Following 30 days of co-culture , we prepared cell lysates and assessed RhCMV-SIV vector replication on the basis of expression of SIV antigen by western immunoblotting . The α-HCMV TRS1 mouse monoclonal antibody that cross-reacts with RhCMV as well as CyCMV in immunoblots was a kind gift from Dr . Adam Geballe ( University of Washington , Seattle , WA ) . The α-GAPDH antibody was acquired from Santa Cruz Biotechnology ( sc-51906 ) . The α-RhCMV IE1 ( 2A1 . 2 ) , α-RhCMV IE2 ( 11A5 . 2 ) and α-RhCMV glycoprotein ( 6H7 . 3 ) antibodies were raised in house by the VGTI Monoclonal Core . Similarly , mouse polyclonal antiserum against RhCMV pp71 was generated by the VGTI Monoclonal Core by immunizing mice with a plasmid encoding for the RhCMV 68–1 pp71 protein . Primary RM or MCM fibroblasts , and telomerized rhesus fibroblasts [69] , were maintained in Dulbecco’s modified Eagle’s medium with 10% fetal bovine serum and penicillin/streptomycin ( DMEM ) at 37°C in humidified air with 5% CO2 . We infected primary RM or MCM fibroblasts with RhCMV 68–1 or RhCMV 68–1 . 2 at an MOI of 0 . 1 ( MOI was calculated on titrations of the RhCMV stocks using primary RM fibroblasts ) and collected supernatant on days 1 , 3 , 5 , and 7 . Limiting dilution plaque assay was used to assess viral titers . As such the dilutions of the viral supernatants were added to immortalized RM fibroblasts and overlaid DMEM-10 containing carboxymethyl-cellulose . Plaque formation was assessed 7 days later by fixing the plates with formalin followed by staining with methylene-blue dye . SIVnef-specific CD8+ T cell lines were generated from SIV-infected RM and MCM as previously described [70] . MHC-matched , primary RM and MCM fibroblasts were infected 48 hours prior to assay set-up with various RhCMV clones expressing a fusion protein of SIV rev , tat , and nef ( RhCMV/rtn ) at an MOI of 1 . These targets were assessed for infection by FACS analysis using an anti-IE2 antibody , and the number of infected targets was normalized across each condition . SIVnef-specific CD8+ T cell lines were mixed with RhCMV/rtn-infected or Nef peptide-pulsed targets at an E:T ratio of 10:1 overnight in Monkey IFN-γ ELISpot plates ( Mabtech , Cincinnati , OH ) . Plates were processed according to the manufacturer’s recommendation and read using an AID plate reader . Data were normalized to the CD8+ T cell IFN-γ release measured after Nef peptide-pulsed fibroblast stimulation . SIV-specific CD4+ and CD8+ T cell responses were measured in mononuclear cell preparations from blood and BAL fluid by flow cytometric intracellular cytokine analysis , as previously described [49] . Briefly , sequential 15-mer peptides ( overlapping by 11 amino acids ) comprising the SIVmac239 Gag , Pol , or Env proteins were used in the presence of co-stimulatory CD28 and CD49d monoclonal antibodies ( BD Biosciences ) . Cells were incubated with antigen and co-stimulatory molecules alone for 1 hour , followed by addition of Brefeldin A ( Sigma-Aldrich ) for an additional 8 hours . Co-stimulation without antigen served as a background control . Cells were then stained with fluorochrome-conjugated monoclonal antibodies , flow cytometric data were collected on a LSR II ( BD Biosciences ) , and data were analyzed using FlowJo software version 10 . 0 . 8 ( Tree Star ) . Response frequencies ( IFNg+/TNFa+ ) were first determined in the overall CD4+ and CD8+ population and then memory corrected ( with memory T cell subset populations delineated on the basis of CD28 and CD95 expression ) . The sequence information for the novel NHP CMV isolates described in this study , RhCMV 19262 ( KX689267 ) , RhCMV 19936 ( KX689268 ) , RhCMV 24514 ( KX689269 ) , CyCMV 31906 ( KX689263 ) , CyCMV 31907 ( KX689264 ) , CyCMV 31908 ( KX689265 ) and CyCMV 31909 ( KX689266 ) have been submitted to GenBank . During our phylogentic analysis , the following DNA and amino acid sequences found in GenBank were used: GPCMV 22122 ( KC503762 , AGE11533 ) , MCMV Smith ( GU305914 , P27172 ) , AoHV-1 S34E ( FJ483970 , AEV80760 ) , SaHV-4 SqSHV ( FJ483967 , AEV80915 ) , CyCMV Ottawa ( JN227533 , AEQ32165 ) , CyCMV Mauritius ( KP796148 , AKT72642 ) , RhCMV 68–1 ( JQ795930 , AFL03576 ) , RhCMV 180 . 92 ( DQ120516 , AAZ80589 ) , SCMV Colburn ( FJ483969 , AEV80601 ) , SCMV GR2715 ( FJ483968 , AEV80365 ) , DrCMV OCOM6-2 ( KR297253 , AKI29779 ) , BaCMV OCOM4-37 ( AC090446 ) , BaCMV OCOM4-52 ( KR351281 , AKG51610 . 1 ) , CCMV Heberling ( AF480884 , AAM00704 ) , PpygCMV 1 ( AAM89279 ) , HCMV TR ( KF021605 , AGL96654 ) and GgorCMV 2 . 1 ( ADY62519 ) . | Nonhuman primates are excellent models of HCMV infection and immunity , and their CMVs closely resemble HCMV in genomic organization and coding potential . Here , we expand the existing models by isolating new full-length , non-human primate CMV strains that more closely resemble primary HCMV isolates . In addition , we identify determinants needed for recombinant clones of RhCMV to infect MCM both in vitro and in vivo . Given their simplified immunogenetics , these results enable the use of MCM to deepen our understanding of CMV immunology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"body",
"fluids",
"immunology",
"vertebrates",
"animals",
"fibroblasts",
"mammals",
"primates",
"urine",
"connective",
"tissue",
"cells",
"genome",
"analysis",
"cytotoxic",
"t",
"cells",
"... | 2016 | Cross-Species Rhesus Cytomegalovirus Infection of Cynomolgus Macaques |
Anti-TNF agents have been in the first line of treatment of various inflammatory diseases such as Rheumatoid Arthritis and Crohn’s Disease , with a number of different biologics being currently in use . A detailed analysis of their effect at transcriptome level has nevertheless been lacking . We herein present a concise analysis of an extended transcriptomics profiling of four different anti-TNF biologics upon treatment of the established hTNFTg ( Tg197 ) mouse model of spontaneous inflammatory polyarthritis . We implement a series of computational analyses that include clustering of differentially expressed genes , functional analysis and random forest classification . Taking advantage of our detailed sample structure , we devise metrics of treatment efficiency that take into account changes in gene expression compared to both the healthy and the diseased state . Our results suggest considerable variability in the capacity of different biologics to modulate gene expression that can be attributed to treatment-specific functional pathways and differential preferences to restore over- or under-expressed genes . Early intervention appears to manage inflammation in a more efficient way but is accompanied by increased effects on a number of genes that are seemingly unrelated to the disease . Administration at an early stage is also lacking in capacity to restore healthy expression levels of under-expressed genes . We record quantifiable differences among anti-TNF biologics in their efficiency to modulate over-expressed genes related to immune and inflammatory pathways . More importantly , we find a subset of the tested substances to have quantitative advantages in addressing deregulation of under-expressed genes involved in pathways related to known RA comorbidities . Our study shows the potential of transcriptomic analyses to identify comprehensive and distinct treatment-specific gene signatures combining disease-related and unrelated genes and proposes a generalized framework for the assessment of drug efficacy , the search of biosimilars and the evaluation of the efficacy of TNF small molecule inhibitors .
In an era of unprecedented accumulation of biomedical data , our understanding of the mechanisms of the development of complex diseases is greatly enabled by the performance of high-throughput experiments and their subsequent analyses at various levels that range from single genes to biological pathways , modules and networks [1 , 2] . Genome-wide transcriptomic profiling has been instrumental in providing accurate representations of the expression programs in homeostasis and disease as well as before and after pharmaceutical interventions [3–7] . Among various pathological conditions , inflammatory diseases such as Rheumatoid Arthritis ( RA ) present great challenges in the understanding of the process through which an initial trigger may lead to generalized and highly variable changes at molecular , cellular and eventually organ level [8] . In this respect , animal models have proven invaluable in the detailed study of these intricate mechanisms and have been the choice of preference in many studies due to particular advantages such as accessibility of material , robustness and standardization [9 , 10] . Recently , an attempt to assess the differential properties of different substances used in the treatment of RA was performed at whole blood transcriptome level of human patients [11] , producing different gene signatures that reflected the mechanistic differences of the tested biologics ( an anti-TNF , an anti IL6-R and an inhibitor of T-cell maturation ) . Among the various therapeutic agents used in the treatment of RA , anti-TNF antibodies have been the primary line of defense since TNF was shown to be a major driver of the disease [12] . In this context , a concise analysis of the effect of the different anti-TNF biologics at transcriptome level has been lacking . The need for investigating the variability in differential patterns of anti-TNF agents [13] has been supported by findings at various levels that include the stratification of RA subtypes [14] , cell-type dependent responses [15] and variability among cells of the same type , namely fibroblast-like synoviocytes [16 , 17] . In this work , we present a computational analysis [18–24] based on a standardized and highly robust protocol involving the administration of four different anti-TNF agents in the hTNFTg ( Tg197 ) humanized mouse model of RA , that has been essential in proving the central role of TNF in the arthritis pathology and its validity as a major therapeutic target [12] . The large number of analyzed profiles and the incorporation of both untreated and healthy samples in our study , enables us to go beyond a simple recording of differentially expressed genes and enriched pathways , which are the usual output of such analyses . Instead , we present a framework for the assessment of drug similarity at various functional levels through the implementation of a combination of gene clustering and state of the art classification methods . Treatment profiles were analyzed at two levels , focusing first on genes that are associated to the disease and then on those that are altered by the treatment even if unchanged in transgenic ( diseased ) animals . In this way we are able to pinpoint a number of functional attributes that are treatment-specific and accordingly devise measures of profile similarity between treatments and the healthy state . Our results suggest subtle , yet significant differences between the different biologics as well as between different intervention timing ( early or late ) . Our work provides a general framework for the comparison of treatment-specific transcriptomes that may assist in a ) the detailed profiling of the gene and functional modules addressed ( or left unaffected ) by a given treatment b ) the multi-level assessment of the treatment’s efficiency in restoring gene expression levels and c ) similarity searches between different pharmaceutical agents . Implementation of our analyses may thus lead to interesting applications in the search of biosimilarity and specify suggestions in the evaluation of small molecule inhibitors .
The hTNFTg ( Tg197 ) human TNF transgenic model develops chronic inflammatory polyarthritis with 100% incidence and with clinical manifestations and histological findings very similar to those of the human disease [12] becoming evident as early as 3 weeks of age with signs progressively worsening as animals age . Animals were treated in a therapeutic regimen from week 6 of age when pathology is already established or in a prophylactic regimen from week 3 of age when pathology is at an early stage . Groups of mice of the same age ( gender balanced ) received either saline or one of the following anti-TNF agents: a ) the chimeric monoclonal antibody infliximab ( Remicade , Janssen Biotech ) , b ) the pegylated antigen binding fragment ( Fab ) certolizumab pegol ( Cimzia , UCB ) c ) he fully human monoclonal antibody adalimumab ( Humira , Abbvie ) and d ) the fusion protein TNFR2/IGg1Fc etanercept ( Enbrel , Pfizer ) . The first three were administered at 10 mg/kg intraperitoneally twice weekly , while etanercept was administered subcutaneously at 30 mg/kg thrice weekly . Apart from a standard therapeutic stage administration , infliximab was also administered with the same dosage but starting at an earlier stage ( 3 weeks ) . During a treatment period of 6 weeks mice were regularly monitored and scored for the progress of the disease pathology and for their overall health status . All therapeutic treatments showed very similar disease progression patterns with an overall in vivo arthritic score dropping steadily from the onset of the intervention , suggesting similar patterns of remission of the hTNFTg arthritis pathology ( Fig 1A ) . Sets of Differentially Expressed ( DE ) genes were defined on the basis of standard thresholds i . e . a |log2FC|≥1 combined with an adjusted p-value≤0 . 05 , against either wild-type ( WT ) or transgenic ( TG ) samples . The cutoffs used , corresponded to values lying beyond two standard deviations ( +/-2σ ) in a symmetrical log2FC distribution and resulted in the selection of <5% of the total genes . We then combined DE genes in various lists depending on the initial comparison undertaken and focused our analyses on the following DE gene subsets: For the initial part of our analysis we focused on the first subset of the 867 TG/WT DE genes , which was used as a disease-associated gene set that allowed us to monitor the functions and pathways that are mainly affected in the diseased animals regardless of pharmaceutical intervention . The more extensive sets of 1064 and 1338 DE genes were used in further analyses of treatment-specific properties , since they included genes whose expression was modulated by the treatment even if they remained unchanged in the diseased animals . They thus represented more inclusive lists comprising genes that are not directly related to the diseased state . Expression patterns of the 867 disease-associated genes in treated and untreated hTNFTg against wild-type samples are shown in Fig 1A . One can see that the DE genes are roughly equally divided in over- ( 404 ) and under-expressed genes ( 463 ) in the diseased versus wild-type state . The slight predominance of under-expressed genes is a distinguishing feature of our analysis compared to previous attempts on human samples , where the lack of normal controls often shifts the balance to over-expressed genes in ratios that vary from 3:1 to 6:1 [25–27] . The large number of under-expressed genes points towards a number of functions that are associated to the diseased state but are not necessarily linked to known , implicated immune and inflammatory pathways . Focusing on particular treatment profiles , we found that the prophylactic intervention with infliximab was the one that most readily restored expression to wild-type levels . This was expected due to the time of intervention at 3 weeks of age and prior to full disease development . On the other hand , in this particular intervention , gene expression changes were in many cases over-compensated , leading to expression levels being reversed , that is genes over-expressed in diseased mice , being under-expressed in treated samples and vice versa . Among the therapeutic interventions , infliximab appears to be the treatment with more direct effects on over-expressed genes , while adalimumab , etanercept and certolizumab pegol present great similarity in terms of expression levels , showing increased restoring potential for a large set of under-expressed genes . Differences between infliximab and the rest of the treatments were primarily quantitative , an observation that is in accordance with a mild advantage of infliximab in the clinical data ( see Fig 1A ) . For specific subsets of genes , restoration to normal ( wild-type ) levels was more extensive with infliximab compared to the other three substances . Based on the expression patterns of treated and untreated mice , we clustered the 867 disease-associated genes in 8 distinct clusters , a number that was indicated by the application of a Silhouette consistency test ( see Materials and methods ) . Of the 8 clusters , 5 contained genes that were under-expressed in disease ( 463 genes ) , while 3 included a total number of 404 over-expressed genes , as may be seen in the coloured side bar of Fig 1B . A more detailed inspection of the differential expression values for each cluster from Fig 1B reveals different patterns of response for the different treatments . Among the three over-expressed gene clusters , the first ( light blue ) shows all treatments sufficient to bring expression down to wild-type levels . The second cluster ( dark green ) contains genes for which prophylactic intervention drives a partial reversal of gene expression levels , while therapeutic treatment is more or less comparable regardless of the administered biologic , with a slight but quantifiable advantage for infliximab . The third cluster ( yellow ) comprises the subset of the most over-expressed disease-associated genes ( compare with Disease state in Fig 1B ) which are , expectedly , partially restored to normal levels by therapeutic interventions but are , interestingly , strongly reversed in the prophylactic one ( see also S1 Fig ) . This reversal may indicate that the subset of the most over-expressed genes is probably not present , or at least not fully formed before later stages of the disease . Thus , it contains genes , which are reversed when the treatment is administered early but not fully addressed at a later stage , an observation which may bear significant implications on the dynamics of the disease . Under-expressed genes are clustered in five groups of variable size . Going from top to bottom ( Fig 1B ) , for two of the five ( Fig 1B , shown in orange and red ) there is a quantitative difference in gene expression restoring potential between adalimumab , etanercept and certolizumab pegol against infliximab . What is also worth mentioning , is that for these clusters infliximab shows limited capacity to modulate gene expression at prophylactic intervention . This may be indicative of different expression dynamics between over- and under-expressed genes , with the first being deregulated earlier and thus addressed at a prophylactic stage while the latter being associated with a more delayed onset . A third under-expressed gene cluster ( dark blue ) comprises 61 genes for which gene expression is partially restored with all treatments , while a fourth , shown in brown at the bottom of Fig 1B shows partial resetting of gene expression levels with therapeutic interventions and a reversal ( from under- to over-expressed ) in the case of the prophylactic infliximab treatment . Lastly , a small cluster of only 12 genes ( Fig 1B pink ) , contains the genes with the most acute under-expression patterns , which seem to be preferentially addressed by infliximab at prophylactic intervention but not by the rest of the treatments . The small number of these genes ( 12 ) does not allow for an enrichment analysis but in general they correspond to genes associated with collagen and myofibrils ( Col10a1 , Glt25d2 , Xirp2 , Myl2 and Clec3a ) innate immunity ( C7 , Cytl1 , Vsig4 ) , the phospholpases Pla1a and Pla2g2a , the serine peptidase Htra4 and the developmental protein Hhip ( heghehog interacting protein ) . In order to gain insight on the functional characteristics of the underlying genes we performed a functional enrichment analysis individually for each cluster . Functional enrichments were calculated with the implementation of gProfiler [20] as described in Methods and a summary of the most enriched terms at the levels of Gene Ontology , KEGG pathways and Transcription Factor targets is shown in Fig 2A and 2B . Two main observations stem from the enrichments of over- and under-expressed gene clusters respectively . On one hand , terms related to the inflammatory and immune responses , including cytokine and chemokine signaling as well as infectious pathways and the associated transcription factors are prominent in all three over-expressed gene clusters ( Fig 2A ) . On the other hand , there is a characteristic over-representation of functional terms related to muscle and heart functions and associated diseases , as well as metabolic and developmental pathways among the under-expressed genes ( Fig 2B ) . Heart failure and related cardiovascular diseases are known comorbidities of Rheumatoid Arthritis patients [28–30] , while we recently showed this to be the case for the animal model under study [31] It is thus interesting to see that associated functions are linked with the under-expression of certain genes , which , as noted above , have been largely overlooked in human studies . In this respect , the aforementioned differential capacity of the analyzed anti-TNF agents to effectively restore under-expressed genes , could lead to more detailed insights on treatment-specific efficiency regarding indirect and secondary effects of the disease . As a next step we wanted to assess the level of potential for gene expression resetting that may be achieved by each treatment . For this we employed a distance-from-WT calculation ( see Materials and methods ) for all disease-associated genes and for each of the gene expression clusters defined above . Mean absolute log2FC values were calculated for each cluster and for each treatment and the distance from WT was then used as an indication of treatment efficiency . The results are summarized in Fig 1C , where distances from WT are shown in the form of a clustered heatmap . Overall mean distances ( all genes , black ) show small differences between treatments with infliximab lying closer to the healthy state , regardless of time of intervention . When looking into separate clusters one can observe that clusters associated with over-expressed genes , and with immune and inflammatory responses , respond better to infliximab at prophylactic stage . For the same over-expressed clusters , infliximab seems to have a small quantitative advantage at the therapeutic stage , over the rest of the treatments . On the other hand , adalimumab in particular but also etanercept and certolizumab pegol , are more readily restoring gene expression levels of under-expressed gene clusters , an effect that is especially strong for two of the largest clusters ( orange and red ) which are functionally associated with heart-related functions and diseases ( see Fig 2B ) . This similarity of the three agents is also reflected in the treatment clustering , in which they are placed together in a group that is separate from inflixιmab . It thus seems that one particular property discriminating infliximab from the rest of the other three anti-TNF agents is related to its increased efficiency in modulating the expression of inflammatory and immune-related genes that are up-regulated in disease . On the other hand , adalimumab , etanercept and certolizumab pegol show greater capacity in restoring gene expression levels for under-expressed genes associated with RA comorbidities . When assessing pharmaceutical interventions , one of the main aspects we need to address is general changes that are not directly associated with the condition under treatment . In the context of our transcriptomic analyses this meant looking into the genes , whose expression changed between treatment and wild-type profiles but which were not altered in the diseased state , as well as into genes that were changed between treatment and diseased samples in general . In this sense , we may divide differentially expressed genes in three groups: Analysis of the corresponding genes belonging to each category resulted in a total of 1338 genes that were unequally distributed among treatments ( S2 Fig ) . At a first-level quantitative assessment based on the number of genes , infliximab performs better in terms of restoration to wild-type levels , at both prophylactic and therapeutic interventions , with the numbers of non-restored genes being only 3 and 25 ( out of a total of 867 ) respectively , compared to significantly higher numbers of non-restored genes for adalimumab ( 92 ) , etanercept ( 106 ) and certolizumab pegol ( 175 ) . When assessing the treatments through their distance from WT for this extended set of genes , it is infliximab at therapeutic intervention which now shows the highest efficiency ( mean distance = 0 . 47 ) followed by the same substance when administered at prophylactic stage ( mean distance = 0 . 53 ) . Among the three remaining biologics differences are becoming clearer with adalimumab performing considerably better ( mean distance = 0 . 61 compared to etanercept’s 0 . 72 and certolizumab’s 0 . 78 ) ( S3 and S4 Figs ) . It may be noted here that , variability in human patients’ response to adalimumab treatment has been attributed to differential levels of cytokine expression in the synovium [32] , which may reflect the quantitative effect compared to infliximab that we are observing . At this first , purely quantitative level , our results are indicative of a more direct response at gene expression level for infliximab ( particularly at prophylactic intervention ) and adalimumab compared to etanercept and certolizumab pegol . When one looks at the altered genes , a time effect appears to become important , with infliximab at prophylactic intervention showing a greater number of genes ( 170 ) which is comparative to certolizumab ( 175 ) and etanercept ( 217 ) , the latter being the therapeutic treatment with the highest number of non-disease associated modulated genes . Indeed , the time of intervention appears to be crucial since the same substance ( infliximab ) when administered at a later stage ( 6w ) has practically no altered genes , suggesting maximal specificity . Overall , infliximab and adalimumab at a therapeutic stage of intervention show the greater number of restored genes combined with a small number of altered genes and thus appear , at this level , to reflect the most effective gene expression modulation patterns . In order to gain more detailed insight on how the response to treatment may be compared to the diseased state , we plotted gene expression values for the three gene categories ( Restored: green , not Restored: red , Altered: blue ) in two dimensional scatterplots shown in Fig 3A . Here , the expression of each gene in disease is shown on the horizontal axis and the corresponding value for the treated samples is shown on the vertical one . In this respect , a good overall response will have points distributed on a horizontal line around 0 ( no change compared to wild-type ) , while points that fall away from the horizontal baseline will represent genes that are either non-responsive to treatment ( red ) or genes that are altered ( blue ) . The diagonal dashed line corresponds to identical values in treatment and disease , therefore linear trendlines , based on the total number of genes , are representative of the overall response profile of the treatment . Slopes greatly deviate from 1 ( diagonal ) , with infliximab again showing the smallest slope values , closer to the desired horizontal . Infliximab at prophylactic intervention shows the smallest slope value , but this is mostly due to the small deviation of the altered genes . A close inspection of Fig 2A shows that altered genes ( blue ) for infliximab at prophylactic intervention fall on the opposite side of the trendline , which is indicative of gene expression reversals . One may observe this in Fig 3B , where ( on the right ) the gene expression values for the genes altered in this particular treatment show a strong reversal of expression levels against the diseased animals . Thus , a significant number of genes seem to be affected in a very acute way when intervention takes place early , as may be seen with a direct comparison to the infliximab therapeutic treatment ( Fig 3B , right ) . We next performed a detailed analysis of partial overlaps of not restored ( Fig 3C ) and altered genes ( Fig 3D ) . Genes not restored to normal levels do not appear to be treatment-specific , as may be seen in extensive overlaps among all treatments ( Fig 3C , layer of black dots spanning most conditions ) . Certolizumab pegol and etanercept in particular , share the highest number of not restored genes ( 88 genes in total , 20 of which are specific between them ) . Infliximab has 25 genes whose levels are not restored , 21 of which are shared with adalimumab . As may be seen in Fig 3A , these genes ( red dots ) are primarily over-expressed in disease and even though their deregulation is ameliorated upon treatment they remain quantitatively active , above the thresholds that qualify them as differentially expressed . This quantitative effect may explain the functional enrichments of not restored genes shown in the table accompanying Fig 3C . With the exception of infliximab , whose 25 not restored genes are enriched in functions related to the extracellular matrix and its degradation , the other three treatments show enrichments in immune , inflammatory and infectious pathways ( see also S5 Fig ) . Even though this seems , at first , counter-intuitive it is supported by the results in Fig 3A and 3B where one can see that there is a quantitative lag in restoring the expression of disease-associated genes . Thus , even though treatments may show similar macroscopic effects ( Fig 1A ) , they have quantifiable differences at the molecular level ( S4 Fig ) , an observation that highlights the importance of -omics approaches in the analysis of detailed treatment profiles . An identical gene overlap and functional analysis was carried out for altered genes ( Fig 3D ) . In the case of altered genes , effects appear to be largely treatment-specific , with most of the altered genes belonging to particular treatments ( compare layers of dots in 3C and 3D ) . As noted earlier , the most striking observation is the significant number of altered genes in the case of infliximab prophylactic intervention ( 170 ) . Of these , 164 are exclusive to the particular treatment , which is indicative that they are largely associated to the age and the different dynamics of gene expression of younger mice that have not yet manifested the disease in its full proportions . These differences are also reflected in the functional analysis where terms related to development , cell motility and fatty acid and lipid metabolism are the ones primarily enriched . What is particularly interesting for altered genes of the infliximab prophylactic treatment , is their expression levels , which show a strong reversal compared to the disease state ( Fig 3B , right ) . This is yet another indication of the strong effects of the prophylactic intervention , effects that extend beyond the disease-associated gene set . Altered genes of adalimumab , etanercept and certolizumab are also largely treatment-specific ( Fig 3B , right ) . Etanercept and certolizumab share a significant percentage due to the overall large numbers of altered genes ( 217 and 175 respectively ) . At the functional level , however , differences are also apparent , with certolizumab’s genes being enriched in structural proteins and oxidative metabolism , etanercept’s being very strongly associated with skin functions and differentiation and adalimumab having only a handful of enriched pathways related to structural proteins . Together , the analysis of not restored and altered genes reveals differences at both qualitative ( gene lists and functions ) and quantitative ( extent or reversal of gene expression levels ) . Therefore , accurate representations of treatment regimens at transcriptomic level require the analysis of such inclusive gene sets . We set out to build a classification model that would preferably discriminate between the treatments and the diseased samples , while at the same time , provide us with insight on the most important genes and pathways . We chose a Random Forest approach since it combines both such characteristics and implemented it on the total number of profiles , including wild-type and transgenic [24] . The best out of 1000 models ( see Materials and methods ) gave a perfect discrimination between wild-type and transgenic profiles ( which was expected given their considerable differences ) but also led to a significant improvement in the classification of the different treatments . The results of the Random Forest classification are visualized in Fig 4 first as PCA plots based on the total of 1338 differentially expressed genes ( Fig 4A ) as opposed to using only the top 100 most important genes based on the Random Forest classification model ( Fig 4B ) . ( Notice that given the large number of analyzed features , PCA is used here as an indicative representation of the capacity of the Random Forest model to discriminate between the different conditions ) . A common characteristic of both PCA analyses is related to a greater variability of transgenic profiles , compared to wild-type and treated samples . Such variability is widely reported in humans and is considered a hallmark of complex diseases [3] , however , an even greater number of samples will be required in order to conclude that this variability may be accurately reflected and quantified through our approach . In terms of classification efficiency ( Fig 4C ) there is a near perfect classification of wild-type and diseased samples , as expected given that the genes under consideration are predominantly differentially expressed between them . Among the various treatments infliximab at both stages of intervention clusters closely to adalimumab . On the other hand , etanercept and certolizumab pegol are classified together in a separate subgroup . When looking closer at the genes predicted by the model to act as the most important classifiers we find a number of genes involved in functions related to the cell adhesion and the extracellular matrix ( Fig 4D ) . The lack of the dominant infectious , immune and inflammatory pathways is expected since the scope of the classification is to define the genes that would better discriminate between all profiles , including the 5 different treatment regimens . Thus , the genes that achieve the best classification are mostly revealing of inter-treatment differences as may be seen in their expression profiles . When looking into gene expression levels of the top 100 most important genes ( Fig 5A ) one can see a predominance of genes that are under-expressed in diseased animals since it is among them that the greater variability among treatment becomes manifest . Fig 4A and 4B suggest that the main premise of the random forest classifier is a distinction between the diseased and wild-type states . As this is likely to be obscuring a better discrimination between treatments , we employed an identical classification approach on a restricted set of samples that included only the treated animals . The results of this classification ( S6 Fig ) show a marked differentiation of infliximab profiles compared to the three other substances , which now show partial but not complete overlaps . The most important genes in the classification are again primarily associated with functions related to the extracellular matrix , development and cell adhesion ( S7 and S8 Figs ) . Inspection of the gene expression levels of the top 300 most important genes shows that classification is based predominantly on two properties: a ) the tendency of infliximab at prophylactic intervention to drive a reversal of over-expressed genes and b ) the generally limited capacity of certolizumab pegol to restore over-expressed genes ( S8 Fig ) . Together this first level Random Forest analysis suggests that even though significant differences between treatments can be identified , the most important features are somehow lacking descriptive power . That is the genes that best classify the samples were only partially reflecting the system under study . In order to further dissect these differences , we next employed a Random Forest analysis at functional level . We applied an identical Random Forest approach at the functional level , starting from a set of differentially expressed genes , mapping them to the corresponding functional categories and then using the mean expression value as input feature for the model ( see Materials and methods ) . By applying the same strategy as the one described above , we were able to define the most important GO , KEGG Pathway and Transcription Factor ( TF ) categories in the discrimination of treatment profiles . The results from the best model are summarized in Fig 5B where the top 30 most important KEGG Pathways are shown for the best model applied on the full dataset comprising all samples ( including healthy and diseased states ) . When looking at functional level , both pathways ( Fig 5B ) and transcription factor features ( S9 Fig ) , accurately reflect the system under study with major inflammatory and immune disease pathways and related transcription factors being among the most important predictive characteristics of the model . This indicates that functional enrichment analyses may confer additional , if not superior insights in the system under study , compared to the more detailed inspection of particular genes . When classifying treatments in the absence of healthy and diseased profiles , thus having removed strictly disease-associated pathways , a first interesting observation is the role of metabolic pathways , which appear to be dominant both at the level of KEGG Pathways ( Fig 5C ) and Transcription Factor enrichments ( S9 Fig ) . Nitrogen metabolism in particular , including nucleobase and aminoacid rank among the top most important pathways , while , at the level of TF , immune-response transcription factors such as Nfkb have given their place to metabolic , growth and developmental regulators such as Pparg , Atf1 and Hoxc8 ( S9 Fig ) . In an attempt to capture a more complete picture of treatment efficiency we calculated mean distances from both wild-type and transgenic samples for each set of profiles . In principle , profiles need to be far from the transgenic but close to the wild-type state , but in the actual data a variety of intermediate profiles may be observed . In order to monitor the range of these responses we have devised a simple two-dimensional approach , that aims to capture , describe and visualize differences between treatment profiles at functional level . The approach is generic which means that it can be readily applied to any gene categorization ( see Materials and methods for details ) but is , herein , restricted to the level of already discussed GO , KEGG Pathway and TF functional categories . We went on to interpret and visualize these enrichments in the following way: For any given functional category we have calculated a mean distance of expression of each treatment ( taking into account only DE genes ) against both wild-type and transgenic samples . Then , for a selected list of such functional categories , we plot the distances from both profiles ( wild-type and transgenic ) in the form of density scatterplots as those shown in Fig 6A . These represent the density of DE genes belonging to the selected functional categories in two dimensions , in which the vertical and horizontal displacements from 0 correspond to the distances from diseased and healthy samples respectively . In this sense , efficient responses are represented by clouds lying around the vertical axis x = 0 , with minimal horizontal ( low adverse effects ) and maximal vertical ( high response to disease ) values ( also see Materials and methods ) . Fig 6A shows the combined density plots for the top 30 most important KEGG pathways described by the best model in the Random Forest classification . Infliximab at prophylactic intervention stands out with the most preferable pattern , represented as a cloud along the vertical axis and with a small displacement towards negative values for the distance from wild-type . Differences between the therapeutic interventions may be qualitatively observed in the shape of the density plot . Starting from infiximab and down to certolizumab pegol , the contours flatten out along the horizontal axis pointing to increasing levels of distance from wild-type . All therapeutic interventions share similar patterns with vertical displacements ( distance from transgenic ) systematically greater than the horizontal ones ( distance from wild-type ) but with considerable distances from both diseased and healthy states , represented as diagonally diffuse clouds . This approach may be used for different functional categorizations in order to capture particular characteristics of the response , as may be seen in the case of GO terms or Transcription Factors ( S10 and S11 Figs ) . Again , starting from infliximab the density plots show a gradually increasing displacement along the x-axis , which is representative of insufficient restoration of gene expression to healthy , wild-type levels . Some interesting features of this analysis are related to the persistence of the retinoic acid receptors ( S11 Fig ) whose targets appear to be invariably under-expressed in transgenic but over-compensated in the treatments . Differences between treatments may be primarily attributed to metabolic regulators such as Pparg and Ppara as well as Atf1 , and Atf2 , which have been reported as being selectively under-expressed in RA synovial extracts even when compared to osteoarthritis ( OA ) controls [33] . Two-dimensional density plots may provide a helpful framework for the visualization of the treatment profiles but are not easily quantifiable . Using , the 2D profiles as starting point , we devised a simple measure of response efficiency that aims to capture a combination of a treatment’s potential in a ) restoring expression levels that are changed in disease while b ) leaving wild-type , healthy levels unaltered . We calculated such an efficiency measure as the log10-transformed ratio of the absolute transgenic over the absolute wild-type distance ( see Materials and methods for details ) , for various subsets of functional categories . The results for the most important KEGG pathways are shown in Fig 6B , where one may observe particular tendencies of the different responses in great detail . The overall greater efficiency of the prophylactic intervention is again apparent , however there is also a greater consistency in the pattern of infliximab at therapeutic intervention with very few negative scores ( which correspond to high frequency of gene expression reversals ) . Perhaps the most interesting observation from the efficiency analysis comes from the bottom of Fig 6B where one can see increased positive scores for adalimumab , etanercept and , to a lesser extent , certolizumab pegol for a number of functions related to heart disease as well as two metabolic signaling pathways ( Adipocytokine and Insulin signaling ) , for which infliximab shows lower efficiency at both stages of intervention . These results are further supported by more extensive analyses for broader functional sets ( Fig 6C ) as well as at the level of transcription factor enrichments ( S12 Fig ) . In fact , a recurring observation from various points of our analysis points towards a general pattern for adalimumab and etanercept targeting pathways associated with known RA comorbidities more effectively than infliximab , compared to a less pronounced response against key inflammatory processes which are more readily addressed by the latter ( Fig 7 ) .
The development of anti-TNF therapies has been a milestone in the treatment of rheumatoid arthritis . Currently there are 5 different biologics ( infliximab , adalimumab , etanercept , golimumab , certolizumab pegol ) in the market while novel biologics or biosimilars have also been developed or are under development . Although all of them target the same molecule , they are different in their molecular structures which range from a fusion protein ( etanercept ) , to human ( adalimumab ) and chimeric ( infliximab ) mAbs and a PEGylated Fab fragment ( certolizumab pegol ) . Such structural differences may translate to the differences these agents exhibit in antibody-depended cell-mediated cytotoxicity , complement-depended cytotoxicity , capacity to induce apoptosis , ability to neutralize membrane bound TNF or differential inhibition of TNFR1 and TNFR2 signaling [34 , 35] . These properties , together with the specific pharmacokinetic and pharmacodynamic profile of each anti-TNF biologic , contribute in shaping their clinical performance profile including their efficacy and safety parameters such as their immunogenicity , as well as the extent to which each specific biologic may affect comorbid pathologies , allergic responses and host defense mechanisms [34 , 36–38] . However , comparisons of the different agents exist only in the clinical setting that does not allow head to head or molecular comparisons . With this study we address the need for such comparisons at the mechanistic/molecular level by using an established arthritis model widely used for the preclinical evaluation of anti-TNF therapeutics [12] . To this end we have developed a generalized framework to address differential gene expression in transcriptomic profiles obtained under disease and treatment with 4 different biologics . In the context of RA , we analyze an extended dataset , which consists of a large number of biological replicates and includes both healthy and untreated animals . This enables us to define gene expression changes against both states and thus achieve considerable insight into the way each treatment modulates expression levels . The detail with which the profiles are analyzed allowed us to reveal , previously unreported characteristics related to gene under-expression and off-target responses as well as to pinpoint particular functional attributes that appear , to a great extent , to be treatment-specific . Thus , on one hand , we bring forward hitherto unreported properties of anti-TNF biologics in the context of an established animal model of inflammatory polyarthritis , while , on the other , we describe a concise set of computational analyses for transcriptomic analyses of drug interventions . From the computational analysis point of view , we show that transcriptomics may capture significant differences between anti-TNF treatments that remain largely unobserved at a macroscopic level . Thus , while small differences may be recorded in clinical readouts such as the disease activity scores ( which are by definition coarse and subject to noise ) , more pronounced variability at both quantitative and qualitative levels is revealed through a detailed dissection of gene expression and functional enrichment data , like the one we present in this work . From a computational analysis viewpoint , one particularly interesting aspect is the greater descriptive capacity achieved from the analysis of broader functional terms compared to the “granularity” of genes . As suggested by the last part of our analyses classification at pathway levels provides a more accurate representations of the system under study . These observations are of particular interest when it comes to considering approaches of alignment between animal models and the human condition , as they are suggestive that data integration at hierarchically higher functional levels provides more accurate representations of the conditions under study . Similar systems approaches in human samples , albeit at a smaller scale have shown the increased descriptive power of pathways and modules instead of simple gene signatures [39 , 40] . When focusing into the system under study , the response of inflamed synovial tissue to anti-TNF intervention , the results presented in this work point out to a set of functions and terms as immune response , cell communication , cell cycle and signaling already reported for human patients treated with anti-TNF biologics [32 , 41] as well as a number of largely overlooked features regarding anti-TNF therapeutic approaches with potentially important implications for the human condition . A first point has to do with the importance of under-expressed genes . Most published works in both humans and animal models focus on over-expressed genes as they are clearly enriched in inflammatory and immune response pathways , targeted by anti-TNF agents . Nevertheless , our data show that more than half of the differentially expressed genes have decreased expression levels compared to healthy controls . These are interestingly enriched in functions that are not directly related to inflammation , such as those associated with heart and muscle development and related diseases as well as secondary metabolic pathways . Such functions are not expected to be directly addressed by anti-TNF action . We find , however , that they are differentially modulated by the substances analyzed , an observation that points to interesting treatment-specific characteristics . Another interesting aspect regarding under-expressed genes is that they represent a subset of genes and functions for which a prophylactic intervention fails to efficiently restore expression levels . This may be indicative of a time-dependent effect under which an initial wave of over-expression of inflammation genes is followed by a late onset under-expression , which thus escapes the early intervention . The time of intervention is also shown to be important through our analyses of altered genes . Early , prophylactic intervention schemes are unlikely to form part of treatment regimens in humans but are , nonetheless , valuable in the context of animal models as they reveal certain aspects of disease development , that cannot be monitored otherwise . The reduced potential of early intervention to address under-expression may be related to a very different gene expression programme that is characterized by low inflammation . High levels of inflammation in human synovial tissue have been shown to be positive predictors of anti-TNF response [42] and so this could partially explain the shortcomings of the prophylactic treatment . Last but not least , a number of observations in our work are potentially interesting in the alignment of the mouse with the human condition . The first is the previously overlooked prominence of under-expression of genes related to known RA commorbidities [31] discussed above . A second point is related to the level at which animal models and human patient samples are to be compared . As shown from our random forest analysis , a representative description of the diseased state is better achieved through functional enrichment analysis at pathway level instead of aiming at the definition of gene lists and signatures . In the past we have shown this superiority of pathway level interpretation for the alignment of mouse to human data at both transcriptome and methylome levels [43] . A third point is related to the well-established variability of human patients in terms of both disease severity and response to therapy [13 , 16–17] . Even though the number of samples in our study is limited , such variability is replicated with disease samples showing broader expression patterns compared to wild-type controls , as is evident in both the PCA analyses and the distributions of expression values . Perhaps the most interesting implication of our study may be that the significant quantitative and qualitative differences we detect between different anti-TNF agents are underlying the well-studied variations in patient response . In all , our approach involving head to head comparisons of different anti-TNF biologics aligns to the limited human data and enables us to capture subtle , or more profound differences between anti-TNF agents and to quantify them through an innovative scheme of efficiency scores . We herein show that transcriptomic analyses represent a valuable means for the study of disease mechanisms and the intricate modes of action of specific treatments . The developed computational pipeline may be easily modified and extended to accommodate comparative analyses of drug similarity or small molecule inhibitor efficacy by quantitatively highlighting treatment-dependent discriminatory characteristics . Moreover , such pipelines might be a tool to support the preferential use of a particular agent or class of agents in specific clinical pathology niches thus driving personalized medicine approaches .
WT and human TNF transgenic mice ( Tg197 ) [12] were bred and maintained in a mixed CBA×C57BL/6J genetic background in the animal facilities of Biomedcode Hellas S . A . under specific pathogen-free conditions . Animals were housed in standard plastic cages with wood chip bedding . The animal facility was under an inverted 12:12-h light/dark cycle at a constant temperature of 22 ± 2 °C and relative humidity of approximately 60% . Food pellets and filtered water were provided ad libitum . Experiments were approved by the BSRC Al . Fleming Institutional Committee of Protocol Evaluation in conjunction with the Veterinary Service Management of the Hellenic Republic Prefecture of Attika according to all current European and national legislation and were performed in accordance with relevant guidelines and regulations . Animals were treated in a therapeutic regimen from week 6 of age or in a prophylactic regimen from week 3 of age . Groups of mice of the same age ( gender balanced ) received either saline , infliximab ( Remicade , Janssen Biotech ) , certolizumab pegol ( Cimzia , UCB ) or adalimumab ( Humira , Abbvie ) administered at 10 mg/kg intraperitoneally twice weekly or etanercept ( Enbrel , Pfizer ) administered subcutaneously at 30 mg/kg thrice weekly . At the end of the treatment period all mice were sacrificed and hind limbs were flash frozen . Total RNA was isolated using Trizol reagent from frozen tissues of wild-type ( healthy ) , huTNFTg ( diseased ) and huTNFTg mice that were treated with one of the 4 different agents in therapeutic or prophylactic regimen as shown in S1 Table . All therapeutic interventions were carried out in 10 biological replicates , while infliximab prophylactic was carried out in 3 . Wild-type and transgenic mice were analyzed in replicates of 10 and 13 respectively for a total sum of 66 profiles . All samples were hybridized on the Affymetrix GeneChip Mouse Gene 2 . 0 ST array . Data analysis was performed using Transcriptome Analysis Console ( TAC 4 . 0 ) Software , Applied Biosystems . CEL files were quantile-normalized with RMA . Log-transformed expression measurements were then converted to gene space by calculating mean probeset values referring to the same genes . This was done in order to minimize the complexity of alternative transcript abundance , which we considered , at this level , to be minimal . Only values from genes that were measured in all samples were finally included in the dataset , which consisted of 18704 common measured genes in all 66 samples ( S1 Data ) . Differential expression analysis is usually calculated against a single background condition that corresponds to a baseline control . We took advantage of our experimental setup and calculated differential expression in the treated samples against both wild-type samples , to quantify gene expression changes versus the native , healthy state , and untreated huTNFTg transgenic samples , to assess changes relative the diseased state . This is a considerable advantage of our approach compared to human studies where the lack of healthy controls often undermines a series of comparisons . Differential expression was calculated as log2Fold-Change ( log2FC ) values from a one-way ANOVA followed by Dunnett’s test for multiple comparisons using either Wild-type ( WT ) or Transgenic ( TG ) samples as control condition . The differential expression analysis was implemented in R and is included in S1 Code . Standard thresholds of |log2FC|≥1 and an adjusted p-value≤0 . 05 were applied for the definition of differentially expressed genes . A combination of clustering methods was employed in the clustering of genes and functional enrichments , as well as in the clustering of treatment efficiency scores ( see below ) . Genes and conditions were clustered with agglomerative hierarchical clustering using Ward’s minimum variance criterion [18] . The optimal number of clusters was defined in all cases , on the basis of a Silhouette consistency analysis [19] . Functional analysis was performed with the use of gProfileR [20] , through its R package implementation and separately at the levels of GO terms ( BP: Biological Process , MF: Molecular Function , CC: Cellular Component ) , KEGG pathways and Transcription Factor targets . Transcriptional regulator targets were incorporated through the repositories of RNEA [21] . Profile similarity/distance was assessed in the form of the mean absolute difference of gene expression as log2FC . Thus , for a given number of N genes the distance between two profiles ( P1 , P2 ) is defined as: d ( P1 , P2 ) =∑i=1N|log2FC ( P1 ) gi−log2FC ( P2 ) gi|N ( 1 ) Profile distance from wild-type samples was used as a proxy for treatment efficiency , calculated for the entire set of differentially expressed genes , as well as for various subsets defined through the clustering approaches described above . Comparison of multiple gene lists leads to complex intersection patterns . We used UpSetR [22] , an R implementation of the UpSet technique [23] to analyze the intersection between various gene lists . UpSetR produces visualizations of complex set intersections in a matrix-based layout , while also providing information on the original set sizes . We employed a Random Forest ( RF ) [24] classification strategy to define subsets of genes and functional categories that best discriminate healthy from diseased profiles as well between various treatments . Random Forests were implemented through the randomForest R package . Starting from the complete dataset we used a 70/30% split for training and test sets respectively and built 1000 RF models with 500 trees each , using 10 variables at each split . From the 1000 RF models we chose the one with the lowest out-of-bag error rate and obtained the variables with the greatest importance on the basis of the higher Mean Gini Decrease ( MGD ) . Arbitrary thresholds for the top most important predictors were applied depending on the downstream analyses ( e . g . 30 or 50 for representation reasons , 100 for Principal Component Analysis ) . Random Forest classification was implemented in two different datasets . One with all samples , that included wild-type and untreated transgenic samples and one that excluded them , focusing only on the treatment profiles . We used the first to observe broader differences between healthy and diseased states and the second to provide us with a more detailed view of the treatment-specific characteristics . Two-dimensional analysis of gene expression was calculated as a combination of differential expression values against wild-type ( healthy ) and transgenic ( diseased ) mice profiles . The two-dimensional approach was aggregated at functional level . For each functional category we obtained the DE genes that belonged to that category and then , for each treatment , we calculated the mean log2FC value for this subset of genes versus both wild-type and transgenic samples . This pair of values was then used in two-dimensional representations of treatment profiles and in the calculation of treatment efficiency scores . The two-dimensional density plots represent the landscape of response of a given treatment for a certain subset of genes or functions . They are formed through the aggregation of pairs of distances ( from wild-type and transgenic ) for functional categories specified by the experimenter or derived from previous analyses ( such as the classification schemes described above ) . Each treatment landscape is thus visualized as a contour cloud , the shape and size of which is representative of its efficiency . Displacement along the transgenic ( vertical ) axis corresponds to desirable distances from the diseased state , while displacement from the wild-type ( horizontal ) axis is typical of undesirable effects that place the treatment at a distance from the healthy condition . The relative amplitudes of this data cloud may be quantified in the form of efficiency scores . These are calculated as the log10-based values of the ratios of mean absolute gene expression values of DE against transgenic ( TG ) over wild-type ( WT ) profiles , according to the following formula: E ( P , f ) =log10|log2FC ( Tg ) gi¯||log2FC ( Wt ) gi¯| ( 2 ) where E ( P , f ) is the efficiency score of profile ( treatment ) P for the functional category f , which contains N differentially expressed genes . In ( 2 ) , gi corresponds to a set of N genes belonging to a given category , over which the mean absolute log-fold-change is calculated . High efficiency scores are thus obtained for functional groups with genes having large absolute changes when compared against transgenic , and low when compared to wild-type profiles . All analyses were performed in the R environment with the combination of custom scripts and available libraries . Annotated code is provided in a R Mardown file as S1 Code and the processed data files , required for the full replication of our analysis are provided in one compressed folder as S1 Data . Animal experiments were approved by the Veterinary Service Management of the Hellenic Republic Prefecture of eastern Attika ( Approval license Protocol No . 2478 , 17/01/2011 ) . All processed data generated or analysed during this study are included in this published article . Processed files are provided in a single zipped folder ( S1 Data ) . The code for the analysis is also provided as a R Markdown file ( S1 Code ) . | A number of anti-TNF drugs are being used in the treatment of inflammatory autoimmune diseases , such as Rheumatoid Arthritis and Crohn’s Disease . Despite their wide use there has been , to date , no detailed analysis of their effect on the affected tissues at a transcriptome level . In this work we applied four different anti-TNF drugs on an established mouse model of inflammatory polyarthritis and collected a large number of independent biological replicates from the synovial tissue of healthy , diseased and treated animals . We then applied a series of bioinformatics analyses in order to define the sets of genes , biological pathways and functions that are affected in the diseased animals and modulated by each of the different treatments . Our dataset allowed us to focus on previously overlooked aspects of gene regulation . We found that the majority of differentially expressed genes in disease are under-expressed and that they are also associated with functions related to Rheumatoid Arthritis comorbidities such as cardiovascular disease . We were also able to define gene and pathway subsets that are not changed in the disease but are , nonetheless , altered under various treatments and to use these subsets in drug classification and assessment . Through the application of machine learning approaches we created quantitative efficiency profiles for the tested drugs , which showed some to be more efficiently addressing changes in the inflammatory pathways , while others being quantitatively superior in restoring gene expression changes associated to disease comorbidities . We thus , propose a concise computational pipeline that may be used in the assessment of drug efficacy and biosimilarity and which may form the basis of evaluation protocols for small molecule TNF inhibitors . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [] | 2019 | An integrative transcriptome analysis framework for drug efficacy and similarity reveals drug-specific signatures of anti-TNF treatment in a mouse model of inflammatory polyarthritis |
Pathogens have evolved sophisticated mechanisms to evade detection and destruction by the host immune system . Large DNA viruses encode homologues of chemokines and their receptors , as well as chemokine-binding proteins ( CKBPs ) to modulate the chemokine network in host response . The SECRET domain ( smallpox virus-encoded chemokine receptor ) represents a new family of viral CKBPs that binds a subset of chemokines from different classes to inhibit their activities , either independently or fused with viral tumor necrosis factor receptors ( vTNFRs ) . Here we present the crystal structures of the SECRET domain of vTNFR CrmD encoded by ectromelia virus and its complex with chemokine CX3CL1 . The SECRET domain adopts a β-sandwich fold and utilizes its β-sheet I surface to interact with CX3CL1 , representing a new chemokine-binding manner of viral CKBPs . Structure-based mutagenesis and biochemical analysis identified important basic residues in the 40s loop of CX3CL1 for the interaction . Mutation of corresponding acidic residues in the SECRET domain also affected the binding for other chemokines , indicating that the SECRET domain binds different chemokines in a similar manner . We further showed that heparin inhibited the binding of CX3CL1 by the SECRET domain and the SECRET domain inhibited RAW264 . 7 cell migration induced by CX3CL1 . These results together shed light on the structural basis for the SECRET domain to inhibit chemokine activities by interfering with both chemokine-GAG and chemokine-receptor interactions .
Chemokines orchestrate leukocyte migration during immune surveillance , inflammation , and development [1] , [2] , [3] , [4] . They comprise a large family ( ∼50 ) of small proteins ( ∼7–14 KD ) that are classified into four classes ( C , CC , CXC , and CX3C , where X is any residue ) based on the spacing of conserved cysteine residues at the N-terminus [5] . The CC and CXC classes are by far the largest groups of chemokines , whereas the C class consists of two members ( XCL1 and XCL2 ) and the CX3C class contains only one member ( CX3CL1 ) . All chemokines share a remarkably similar structural fold , consisting of an extended N-terminus , an antiparallel three-stranded β-sheet and a C-terminal helix [6] . Chemokines exert their biological activities through binding with their cognate G protein-coupled receptors expressed on the surface of leukocytes , as well as binding with endothelial or matrix glycosaminoglycans ( GAGs ) to form chemokine gradients along which cells travel across endothelium and into tissues [6] . The molecular basis of chemokine-GAG and chemokine-receptor interactions has not been well understood [6] , [7] . It has been suggested that the basic residues ( typically Arg and Lys ) involved in GAG interaction are more or less scattered along the polypeptide chain and form four distinct clusters on the surface of chemokines [8] , while the N-termini of all studied chemokines is critical for inducing signaling by their respective receptors [6] . The chemokine network is an important component of host immune response to viral infection [1] , [3] , which is also extensively modulated by viruses especially large DNA viruses to evade host reactions . Poxviruses and herpesviruses encode their own chemokines , chemokine receptors and chemokine-binding proteins ( CKBPs ) [9] , [10] , [11] . The viral CKBPs identified so far are unrelated to any host proteins and exhibit diverse chemokine-binding profiles , reflecting differences in viral tropism and pathogenesis . The viral CC chemokine inhibitor ( vCCI , also called T1/35 kDa ) secreted by several poxviruses including cowpox virus ( CPXV ) , ectromelia virus ( ECTV ) and vaccinia virus ( VACV ) is the most extensively studied , which binds many CC chemokines but not C , CXC , and CX3C chemokines to block chemokine-receptor interaction [12] , [13] , [14] , [15] , [16] . The VACV A41 and ECTV E163 , representative members of another family of poxviral CKBPs , interact with a subset of CC and CXC chemokines to block chemokine-GAG interaction [17] , [18] . Mouse γ-herpesvirus 68 encodes a unique CKBP named as the M3 protein that is able to bind chemokines from the C , CC , CXC , and CX3C classes [19] , [20] . Structural and biochemical studies revealed that M3 disrupts both chemokine-receptor and chemokine-GAG interactions [21] , [22] , [23] . Other viral CKBPs , such as M-T7 from myxoma virus ( MYXV ) , a CKBP from orf virus ( ORFV ) , p21 . 5 from human cytomegalovirus and glycoprotein G from α-herpesviruses , have also been described previously [24] , [25] , [26] , [27] . Four different genes encoding viral tumor necrosis factor receptors ( vTNFRs ) have been identified in poxviruses , consisting of cytokine response modifier B ( CrmB ) , CrmC , CrmD , and CrmE [10] . They contribute to pathogenesis of poxviruses and reflect the complex regulation of TNF-mediated host immune response [28] . In addition to the anti-TNF activity attributed to the N-terminal four cysteine-rich domains ( CRDs ) homologous to host TNF receptors [29] , CrmB and CrmD have anti-chemokine activity attributed to a unique C-terminal extension ( ∼160 aa ) , named as the SECRET domain ( smallpox virus-encoded chemokine receptor ) [30] . Biochemical analysis revealed that the SECRET domain binds a subset of human and mouse CC , CXC and C chemokines , including CCL28 , CCL25 , CCL20 , CXCL12 , CXCL13 , CXCL14 , and XCL1 [30] . The identification of other poxvirus genes encoding homologues with the SECRET domain indicates that the SECRET domain represents a new family of viral CKBPs , which has specific folding to allow its binding with chemokines , either independently or fused with vTNFRs [30] , [31] . A recent report predicted the structural homology of the SECRET domain with CPXV vCCI and VACV A41 and also analyzed its structural differences from vCCI and A41 based on a de novo model [32] . Here we report the crystal structures of the SECRET domain of CrmD encoded by an ECTV strain [33] and the complex of it with chemokine CX3CL1 . These structures , together with biochemical and chemotaxis assays , reveal the structural basis for the SECRET domain to bind chemokines and also shed light on its anti-chemokine structural mechanisms .
The crystal structure of the SECRET domain ( residues S162−D320 ) was determined at a resolution of 1 . 57 Å by using single-wavelength anomalous dispersion ( SAD ) method with a Br-soaked derivative ( Table 1 and Figure S1 in Text S1 ) . There are two SECRET domains ( molecules A and B ) in the asymmetric unit ( Figure 1A ) , related by a non-symmetrical two-fold axis with an r . m . s . d . of 0 . 62 Å for all Cα atoms . Although these two monomers bind each other tightly with a buried surface of ∼1160 Å2 , the size exclusion chromatography revealed that it is monomeric in solution ( Figure S2 in Text S1 ) . The same phenomenon was also observed in the CPXV and ECTV vCCI crystal structures [34] , [35] . Therefore , the SECRET dimer in the asymmetric unit is caused by molecular packing and unlikely has any functional significance . The SECRET domain monomer adopts a β-sandwich fold , consisting of two parallel β-sheets and the connecting loops ( Figure 1B and Figure S3 in Text S1 ) . The β-sheet I consists of five anti-parallel strands 1 , 5 , 6 , 10 and 11 ( Figure 1B and Figure S3 in Text S1 ) . The β-sheet II consists of six strands , which can be further divided into two segments ( antiparallel strands 2 , 3 , 4 and 7; antiparallel strands 8 and 9 ) ( Figure 1B and Figure S3 in Text S1 ) . The β-sheet II outside surface is completely exposed to solvent ( Figure 1B ) , whereas the solvent accessibility of β-sheet I outside surface is limited by a long C-terminal loop after strand 11 surrounding the bottom half of β-sheet I ( Figure 1B ) . A disulfide bond , C180−C317 , further fixes the conformation of this extended loop by connecting it to the 1–2 loop ( Figure 1B and Figure S3 in Text S1 ) . The overall β-sandwich topology of the SECRET domain is similar to that of vCCI and A41 [17] , [34] , [35] , [36] . However , there are several significant differences in the arrangement of certain secondary structure elements , making the CrmD SECRET domain more compact than vCCI and A41 and also directly affecting its binding with chemokines . In the following comparison and description , we use the structure of vCCI from ECTV as the representative member of the vCCI family [35] . The first difference is at the 7–9 loop ( S248−H266 ) in the SECRET domain , corresponding to residues S140 to I168 in vCCI and E113 to M144 in A41 ( Figure 2A and Figure S4 in Text S1 ) . The long S140−I168 loop in vCCI wraps the β-sheet I at the top half , and the long E113−M144 loop in A41 wraps the whole β-sandwich from bottom side ( Figure 2A ) . In collaboration with the conserved C-terminal loop surrounding the bottom half of β-sheet I , these two long loops further limit the solvent exposable surface of β-sheet I in vCCI and A41 , respectively ( Figure 2A ) . The 7–9 loop in the SECRET domain goes up and down at the β-sheet II side of the β-sandwich , and residues S252 to Q254 form the strand 8 in β-sheet II ( Figure 2A and Figure S4 in Text S1 ) . Therefore , it does not limit the solvent exposable surface of β-sheet I in the SECRET domain . The second difference occurs at the 2–3 loop ( I184−S186 ) in the SECRET domain , whose length is nearly the same as that in A41 ( K39−Y40 ) and much shorter than that in vCCI ( S52−P66 ) ( Figure 2A and Figure S4 in Text S1 ) . The third difference occurs at the 6–7 loop , whose length in the SECRET domain ( N227−C238 ) is also much shorter than that in vCCI ( S107−C131 ) and A41 ( S80−C104 ) ( Figure 2A and Figure S4 in Text S1 ) . There is an α-helix in this loop region of vCCI and A41 , which is absent in the SECRET domain ( Figure 2A ) . The electrostatic complementarity plays a critical role in the binding of chemokines by vCCI and A41 [17] , [34] , [35] , [36] . The β-sheet II of vCCI exhibits strong electronegative character . Negative charge patches , including the protruded acidic 2–3 loop ( S52−P66 ) ( Figure 3A and Figure S4 in Text S1 ) , are involved in the interactions with positive charged residues of bound chemokine as revealed in the NMR solution structure of vCCI in complex with chemokine CCL4 [36] . The 2–3 loop ( K39−Y40 ) in A41 is much shorter than that in vCCI , but its β-sheet II also exhibits negative charge patches ( Figure 3B and Figure S4 in Text S1 ) and may contribute to the interaction with bound chemokine [17] . The opposite β-sheet I of vCCI and A41 is comparatively uncharged and electropositive , respectively ( Figure 3A and 3B ) . The SECRET domain exhibits different electrostatic surface by switching the surface charge property as observed in vCCI and A41 . Its β-sheet II has no remarkable electrostatic properties , while the opposite β-sheet I exhibits strong negative charge in the solvent exposable region , contributed by acidic residues D167 , E169 , D228 , D290 , D316 , and E318 ( Figure 3C and Figure S4 in Text S1 ) . The distinct surface charge property of the SECRET domain leads us to speculate that it may bind chemokines in a different manner by using the solvent exposable and negatively charged surface of β-sheet I . To directly elucidate the chemokine binding by the SECRET domain , we reconstituted a complex of the SECRET domain with the chemokine domain of CX3CL1 and determined its structure at a resolution of 2 . 6 Å . The structure was solved by the molecular replacement method using the SECRET domain and CX3CL1 structures as search models , and refined to final Rwork and Rfree factors of 19 . 6% and 25 . 0% , respectively ( Table 1 and Figure S1 in Text S1 ) . In the complex , one SECRET domain monomer binds one CX3CL1 monomer , displaying a 1∶1 stoichiometry ( Figure 4A ) . The chemokine domain of CX3CL1 in the complex adopts the typical chemokine-fold topology , consisting of an extended N-loop ( C8−P20 ) , a short 310 helix ( V21−L23 ) , a 3-stranded anti-parallel β-sheet ( β1 , L24−Q29; β2: I39−T43; β3: R47−A51 ) , a C-terminal helix ( Q56−A69 ) packing against the β-sheet , and the 30s loop ( N30−A38 ) and 40s loop ( R44−H46 ) connecting the strands in the β-sheet ( Figure 4A ) . The N-terminal residues Q1 to K7 and C-terminal residues R74 to G76 are disordered in the structure . The SECRET domain contacts the CX3CL1 with its β-sheet I , burying a surface of ∼530 Å2 ( Figure 4A ) . The SECRET domain contacting residues are from the strands 1 , 5 , and 6 of β-sheet I and the C-terminal extended loop , while the CX3CL1 contacting residues are from the N-loop , 310 helix , 40s loop , and the β3 strand ( Figure 4A ) . The binding interface can be described as a small hydrophobic core surrounded by a large halo of hydrophilic interactions . The hydrophobic core is composed of residues Y212 and F225 from CrmD , and I19 , L23 and F49 from CX3CL1 ( Figure 4B ) . The surrounding hydrophilic interactions are composed of hydrogen bonds and salt-bridges . There is an obvious electrostatic complementarity between the SECRET domain and CX3CL1 at the interface ( Figure 4C ) . The acidic residues D167 , E169 , and D316 from the SECRET domain form salt-bridge interactions with R44 , R47 , and K18 from CX3CL1 , respectively ( Figure 4D ) . To further elucidate the roles of important residues in complex formation , we mutated hydrophobic residues I19A , L23A and F49A and charged residues K18A , R44A , and R47A in CX3CL1 , and measured the binding affinities of these mutants with the SECRET domain using surface plasmon resonance ( SPR ) method . We performed two independent measurements for each protein sample and the results are listed in Table 2 . The SECRET domain interacted with CX3CL1 with an affinity of 0 . 68±0 . 26 µM ( Table 2 and Figure S5 in Text S1 ) The CXC3L1 mutants I19A , L23A , and F49A bound the SECRET domain with affinities of 0 . 96±0 . 32 , 4 . 49±1 . 13 , and 4 . 20±0 . 45 µM , respectively ( Table 2 and Figure S5 in Text S1 ) . The CX3CL1 mutants K18A , R44A , and R47A bound the SECRET domain with affinities of 10 . 9±0 . 6 µM , 16 . 15±0 . 45 µM , and 36 . 45±1 . 95 µM , respectively ( Table 2 and Figure S5 in Text S1 ) . All mutations resulted in the decrease of the binding affinity between the SECRET domain and CX3CL1 . Mutating charged residues K18 , R44 , and R47 in CX3CL1 induced more significant binding affinity decrease than mutating hydrophobic residues I19 , L23 , and F49 , suggesting the importance of the salt-bridge interactions in the complex formation of the SECRET domain with CX3CL1 . The SECRET/CX3CL1 and previous reported vCCI/CCL4 complexes [36] are different in the association manner between CKBP and chemokine , binding interface , and the role of electrostatic complementarity in complex formation . The SECRET domain utilizes its β-sheet I to interact with CX3CL1 , whereas vCCI utilizes its β-sheet II to interact with CCL4 upon complex formation . The vCCI/CCL4 binding interface , burying a total surface of ∼990 Å2 , can be divided into two patches . The patch 1 between the N-loop of CCL4 and vCCI is primarily composed of hydrophobic interactions around CCL4 residue F13 and salt-bridge interactions around CCL4 residue R18 ( Figure S6 in Text S1A ) [36] . These two positions are conserved in CC chemokines and mutation of them dramatically decreased the binding of CC chemokines by vCCI [37] , [38] . The patch 2 is between the basic 40s loop of CCL4 and the extended and acidic 2–3 loop of vCCI ( Figure S6 in Text S1A ) , and the electrostatic complementarity is expected to drive the interactions between them , although basic residues K45 , R46 , and K48 in the 40 s loop of CCL4 were mutated to Ala in the vCCI/CCL4 complex structure ( Figure S6 in Text S1A ) [36] . The smaller SECRET/CX3CL1 interface ( ∼530 Å2 ) is composed one contact patch with a small hydrophobic core and surrounding hydrophilic interactions as described above ( Figure 4B ) . The region from C8 to M15 of the N-loop is far away from the SECRET domain , so the SECRET domain does not utilize critical hydrophobic interactions observed in the vCCI/CCL4 contact patch 1 to bind CX3CL1 . The corresponding position of R18 in CCL4 is K18 in CX3CL1 , which forms salt-bridge interaction with D316 of the SECRET domain to surround the hydrophobic core ( Figure 4D ) and is also important for their binding ( Table 2 ) . Structural superimposition based on bound chemokines revealed that the smaller SECRET/CX3CL1 interface corresponds to contact patch 2 in the vCCI/CCL4 interface ( Figure S6 in Text S1A ) . Although obvious electrostatic complementarity is observed in both the SECRET/CX3CL1 interface ( Figure 4C ) and the contact patch 2 in the vCCI/CCL4 interface ( Figure S6 in Text S1A ) , the electrostatic interactions around the basic 40s loop of bound chemokine play a different role in the formation of these two complexes . Residues R44 and R47 of CX3CL1 are critical because mutations at these positions caused respective ∼24-fold and ∼54-fold drop in the binding of CX3CL1 by the SECRET domain ( Table 2 ) . In contrast , a triple mutant of CCL4 ( K45A/R46A/K48A ) had nearly the same binding affinity as wild type CCL4 , as determined by ELISA method [36] . In CCL2 , the K49A mutation even increased its binding affinity with vCCI [37] , [38] . Therefore , the contact patch 2 around the 40s loop of chemokines might contribute to chemokine binding of vCCI by providing an electronegative platform to recruit different CC chemokines , while the conserved hydrophobic and hydrophilic interactions around the N-loop of CC chemokines in the contact patch 1 determine the high affinity binding of CC chemokines by vCCI . The contact patch around the basic 40s loop of CX3CL1 has dual roles , not only helping the recruitment of a subset of chemokines from different classes by the SECRET domain , but also providing critical interactions for the complex formation . Parasites , such as blood-sucking ticks and Schistosoma mansoni , also secrete CKBPs with anti-inflammatory activities [39] , [40] , [41] . Evasin , a new family of CKBPs encoded by ticks , comprises four members that may help inhibit chemokine-mediated host innate immune responses [40] . In contrast to most of the viral CKBPs , Evasin-1 is very restrictive by only binding CCL3 , CCL4 and CCL18 [40] . The Evasin-1 adopts a novel fold and it interacts with bound CCL3 by primarily contacting its N-loop region , as revealed in the Evasin-1/CCL3 complex structure ( Figure S6 in Text S1B ) [42] . Structural superimposition based on bound chemokines revealed that the SECRET-binding and Evasin-1-binding epitopes on chemokines are distinct with little overlap ( Figure S6 in Text S1B ) . Therefore , the SECRET domain and Evasin-1 are different in the chemokine-binding manner . The M3 encoded by murine γ-herpesvirus 68 functions as a dimer in solution , in contrast to other monomeric poxviral CKBPs . The two M3 monomers are arranged in a “head-to-tail” manner , each monomer consisting of the N-terminal domain ( NTD ) and C-terminal domain ( CTD ) ( Figure S6 in Text S1C ) [21] . Unlike the SECRET/CX3CL1 complex in 1∶1 stoichiometry , the M3 dimer utilizes the NTD of one monomer and the CTD of the other monomer to form two clefts to bind two chemokines , forming a complex in 2∶2 stoichiometry ( Figure S6 in Text S1C ) [21] , [23] . The SECRET domain and the NTD of M3 have positional overlap around the 40s loop of bound chemokine ( Figure S6 in Text S1C ) . The Evasin-1 and the CTD of M3 have positional overlap around the N-loop of bound chemokine ( Figure S6 in Text S1C ) . Therefore , M3 seems to combine different chemokine-binding manners of the SECRET domain and Evasin-1 by utilizing both NTD and CTD in the binding of chemokines . We conducted a SPR experiment as reported for other viral CKBPs ( VACV A41 and ECTV E163 ) to test if the SECRET domain can interfere with the interaction of CX3CL1 with GAG [17] , [18] . The CX3CL1 was pre-incubated with various amount of heparin ( sodium salt , molecular weight ∼15 KDa ) , and then injected over the SECRET-coupled sensor chip . Heparin decreased the binding of CX3CL1 by the SECRET domain in a dose-dependent manner ( Figure 5 ) , indicating the overlap of SECRET-binding and GAG-binding sites on CX3CL1 . In our experiment , the concentration of heparin required for the inhibition was much higher than that used to achieve the disruption of chemokine binding by A41 and E163 [17] , [18] . This may be caused by the reported low binding affinity between CX3CL1 and heparin [23] . It has also been shown that the SECRET domain can inhibit the CCL25-mediated Molt4 cell migration , indicating its ability to interfere with binding of CCL25 with its cellular receptors [18] , [30] . We also checked the ability of recombinant CX3CL1 in inducing migration of RAW264 . 7 cell as reported [43] , as well as the ability of the SECRET domain to inhibit cell migration . CX3CL1 induced the migration of RAW264 . 7 cells in a dose-dependent manner , indicated by the decrease of cells remaining in the top well with the increase of CX3CL1 concentration in the bottom well ( Figure 6A ) . Chemokinesis , defined as a random movement of cells in a zero gradient ( equal amounts of starting chemoattractant in both top and bottom wells ) , was very low ( Figure 6A ) . Pre-incubation of CX3CL1 with excessive SECRET domain significantly reduced the CX3CL1-mediated cell migration ( Figure 6B ) . We also expressed and purified the SECRET domain with a triple mutation D167A/E169A/D316A by replacing its acidic residues involved in the critical salt-bridge interactions at the SECRET/CX3CL1 interface . Gel-filtration and circular dichroism ( CD ) spectroscopy profiles indicate that this mutant was properly folded and purified as the wild type protein ( Figure S7 in Text S1 ) . This SECRET domain mutant lost most of the inhibitory ability ( Figure 6B ) . These results together suggest that the SECRET domain is able to interfere with the binding of CX3CL1 with its receptors on cell surface . The measured binding affinity ( ∼0 . 68 µM ) between the SECRET domain and CX3CL1 in our experiment is lower than previous reported binding affinities between the SECRET domain and CCL28 , CCL25 , CXCL12 , CXCL13 , CXCL14 , XCL1 , and CCL20 that are in nM range [30] . This raises the question if the SECRET domain binds other chemokines in a manner similar to that observed in the SECRET/CX3CL1 complex structure . To help answer this question , we examined the binding ability of the SECRET domain D167A/E169A/D316A mutant . Besides CX3CL1 , CCL28 , CCL25 and CXCL12 were chosen because they were the previously reported top three in the binding with the CrmB and CrmD [30] . The SPR analysis showed that the triple mutations in the SECRET domain not only disrupted its binding with CX3CL1 , but also with CCL28 , CCL25 , and CXCL12 to undetectable level ( Figure 7 ) , indicating that the SECRET domain binds different chemokines in a similar manner .
GAG binding plays important roles in the in vivo function of chemokines , including helping immobilize chemokines to form a concentration gradient along which cells can migrate directionally , protecting chemokines from proteolysis , and inducing chemokine oligomerization [44] , [45] . It has been suggested that four distinct basic clusters on the surface of chemokines are major GAG-binding sites for different chemokines [8] . These four clusters all involve residues from the basic 40 loops . Residues on the chemokines important for GAG binding have also been characterized by mutagenesis studies for several chemokines including CCL2 , CCL3 , CCL4 , CCL5 , CXCL8 , CXCL12 , and XCL1 [6] , [7] . Basic residues from the 40s loop participate in the binding of GAG by all studied chemokines except CXCL8 [6] , [7] . The respective structures of CCL5 and CXCL12 with heparin-derived disaccharides also confirmed that the BBXB motif ( where B and X stand for basic and neutral/hydrophobic amino acid ) of the 40s loop participates in GAG binding [46] , [47] . The inhibition of chemokine-GAG interaction by M3 is also attributed to its interaction with the basic 40s loop of bound chemokine by the NTD [23] . These previous results all suggest that the basic 40s loop of chemokines is generally involved in the GAG-binding . In the SECRET/CX3CL1 complex structure , the basic residues R44 and R47 from the 40s loop of CX3CL1 have direct interaction with the SECRET domain . We have also shown that heparin can interfere with the binding of CX3CL1 by the SECRET domain in a dose-dependent manner ( Figure 5 ) , similar to the interference of heparin in the chemokine binding of A41 and E163 [17] , [18] . These data together indicate that the SECRET domain is able to block the chemokine-GAG interaction . The inhibitory ability of the SECRET domain for CCL25 and CX3CL1 induced cell migration indicates that it is able to interfere with the chemokine-receptor interaction . It is generally accepted that the N-termini of chemokines is the key signaling domain , and other residues in the N-loop and core domain can also be critical for the binding with chemokine receptors . For example , the residues 12–17 in the N-loop of CXCL12 were shown to be important for receptor binding [47] , [48] . The N-loop region ( residues 13–20 ) of CC chemokines promotes tight binding to the chemokine receptors [49] , [50] . The vCCI and M3 interfere with the chemokine-receptor interaction by completely blocking the accessibility of the N-loop region of bound chemokine , as revealed in the complex structure of vCCI with CCL4 , and M3 with CCL2 and XCL1 [21] , [23] , [36] . The N-loop of CX3CL1 is not completely blocked by the SECRET domain . The S13 position at the N-terminal part of the N-loop critical for the binding of chemokines by vCCI and M3 is accessible in the SECRET/CX3CL1 complex , but residues T16 , S17 , K18 , and I19 at the C-terminal part of the N-loop region have interaction with the SECRET domain . It suggests that although the SECRET domain does not directly block the most important receptor binding site on CX3CL1 ( i . e . the N-termini and critical hydrophobic residues of the N-loop ) , its binding is still close to the receptor binding site and bring steric hindrance to prevent efficient interaction with the receptor , which would provide a structural basis for the ability of the SECRET domain to inhibit CX3CL1 and CCL25 induced cell migration . The previous study reporting the discovery of the SECRET domain has shown that it is capable of binding CCL28 , CCL25 , CCL20 , CXCL12 , CXCL13 , CXCL14 and XCL1 that are from the CC , CXC , and C classes . We have shown here that it is also able to bind CX3CL1 , the only member in the CX3C class . The ability of previous reported M3 to bind a subset of chemokines from all four classes is attributed to its structural plasticity ( i . e . the structural rearrangement of NTD and CTD ) and the use of flexible loops as primary contact sites for chemokines from different classes [21] , [23] . In comparison , the SECRET domain has a much smaller solvent exposed surface on the relatively rigid β-sheet I to contact chemokines from four different classes , demanding the focus on more common amino acid motifs on chemokines . In the SECRET/CX3CL1 complex structure , critical residues R44 and R47 for the complex formation are from the 40s loop , which can be regarded as hot-spot residues for the interaction . The presence of basic residues in the 40s loop is also found in other chemokines bound by the SECRET domain . The electrostatic complementary between the basic 40s loop of bound chemokine and acidic β-sheet I surface of the SECRET domain would enable the SECRET domain to bind different chemokines , allowing some extent of conformational variation in the 40s loop . There are two questions need to be answered in the future study: ( 1 ) Why is the SECRET domain not able to bind other chemokines also with the presence of basic residues in the 40s loop ? ( 2 ) Why is the binding affinity of the SECRET domain with CX3CL1 lower than with previous reported chemokines ? Sequence alignments of CX3CL1 , CCL28 , CCL25 , CCL20 , CXCL12 , CXCL13 , CXCL14 , and XCL1 did not reveal obvious conserved motifs in the 40s loop ( Figure S8A in Text S1 ) that are absent in chemokines unable to bind the SECRET domain . Previous NMR studies indicated that the flexibility of the N-loop is greater than the flexibility of other regions of chemokines ( excluding the N- and C-termini ) [51] . Only the C-terminal part of the N-loop of CX3CL1 is involved in the interaction with the SECRET domain . Due to the flexibility of the N-loop , it may more extensively participate in the interactions of CCL28 , CCL25 , CCL20 , CXCL12 , CXCL13 , CXCL14 and XCL1 with the SECRET domain , and the chemokine selectivity of the SECRET domain may also reside in the flexible N-loop region . The definite and clear answers to these questions await future structural studies of the SECRET domain with chemokines from C , CC , and CXC classes . Besides CrmB and CrmD , genome analysis also identified other genes encoding SECRET domain containing proteins ( SCPs ) [30] . The reported SCPs that bind to the same set of chemokines as CrmB and CrmD are CPXV V218 ( SCP-1 ) , ECTV E12 ( SCP-2 ) , and ECTV E184 ( SCP-3 ) [30] . The primary sequence of the SECRET domain is much more conserved in CrmB and CrmD than in SCP-1 , SCP-2 , and SCP-3 ( Figure S8B and S8C in Text S1 ) . Among the fifteen residues in the SECRET domain of ECTV CrmD that have contacts with CX3CL1 in complex formation ( Figure 4B and Figure S8B in Text S1 ) , seven of them are strictly conserved in CrmB from VARV and CPXV and CrmD from ECTV and CPXV , including important charged residues D167 and E169 ( Figure S8B in Text S1 ) . Another important charged residue D316 is conserved in CrmD , but is replaced by arginine in CPXV CrmB and serine in VARV CrmB ( Figure S8B in Text S1 ) . Residues interacting with CX3CL1 in the SECRET domain are not highly conserved in SCP-1 , SCP-2 , and SCP-3 ( Figure S8C in Text S1 ) . This indicates that the binding of chemokines by these SCP proteins may be different from the binding by the SECRET domain .
The gene encoding the SECRET domain of CrmD ( residues 162−320 ) was cloned into EcoRI and NcoI restriction sites of the pProEx HTb expression vector . The resulting plasmid was transformed into E . coli BL21 ( DE3 ) competent cells . Three liters of LB media containing 100 µg/ml ampicillin were inoculated and grown to A600 of 0 . 8 and then induced with 0 . 6 mM IPTG . Induced cultures were grown for an additional 4 h at 37°C and harvested by centrifugation for 10 min at 5 , 000 rpm . Cells were resuspended in 25 mM Tris-HCl , 50 mM NaCl , pH 8 . 0 , lysed with sonication and centrifuged for 50 min at 15 , 000 rpm . The SECRET domain was found exclusively in the inclusion bodies . The inclusion bodies were washed three times in wash buffer A ( 25 mM Tris-HCl , 50 mM NaCl , 5 mM EDTA , 5% Triton X-100 , pH 8 . 0 ) and once in wash buffer B ( 25 mM Tris·HCl , 50 mM NaCl , 5 mM EDTA , pH 8 . 0 ) . Washed inclusion bodies were solubilized in 8 M Urea , 50mM DTT and diluted into a refolding buffer ( 25 mM Tris-HCl , 50 mM NaCl , 0 . 2 mM oxidized glutathione , 2 mM reduced glutathione , pH 8 . 0 ) and stirred at 4°C overnight , and then dialyzed against 25 mM Tris-HCl , 50 mM NaCl , pH 8 . 0 . The refolded SECRET domain was bound to HisTrap column , then washed with 25mM Tris-HCl , 50 mM NaCl , 20 mM Imidazole , pH 8 . 0 and eluted with 25 mM Tris-HCl , 50 mM NaCl , 500 mM imidazole , pH 8 . 0 . Fractions containing the SECRET domain were examined by SDS-PAGE gel , pooled and further purified with size exclusion column . The SECRET mutant ( D167A/E169A/D316A ) was expressed and purified by the same method as wild type SECRET domain . To check the SECRET domain is a dimer or monomer in solution , molecular weight standards and the SECRET domain ( 0 . 5 ml , 1 . 5 mg/ml ) were loaded onto Superdex 75 size exclusion column with a flow rate of 0 . 5 ml/min . The gene encoding the chemokine domain of human CX3CL1 ( residues 1–76 ) was cloned into the EcoRI and NcoI restriction sites of the pProEX HTb expression vector . The resulting plasmid was transformed into E . coli BL21 ( DE3 ) competent cells . Three liters LB media containing 100 µg/ml ampicillin were inoculated and grown to A600 of 0 . 8 and then induced with 1 . 0 mM IPTG . Induced cultures were grown for an additional 4 h at 37°C and harvested by centrifugation for 10 min at 5 , 000 rpm . Cells were resuspended in PBS buffer ( pH 7 . 2 ) , lysed by sonication and centrifuged for 50 min at 15 , 000 rpm . CX3CL1 was found exclusively in the inclusion bodies . The inclusion bodies were washed three times in wash buffer A ( 25 mM Tris·HCl , 50 mM NaCl , 5 mM EDTA , 5% Triton X-100 , pH 8 . 0 ) and once in wash buffer B ( 25 mM Tris·HCl , 50 mM NaCl , 5 mM EDTA , pH 8 . 0 ) . Washed inclusion bodies were solubilized in 8 M Urea , 50 mM DTT and diluted into a refolding buffer ( PBS , 0 . 2 mM oxidized glutathione , 2 mM reduced glutathione , pH 7 . 2 ) and stirred at 4°C overnight . Precipitated material was removed by filtration . Refolded protein was bound to a HisTrap column and washed with PBS buffer , 20 mM Imidazole then eluted with PBS buffer , 500 mM Imidazole . Fractions containing CX3CL1 were examined by SDS-PAGE gel , pooled and further purified with size exclusion column . All CX3CL1 mutants were expressed and purified by the same method as wild type CX3CL1 . Purified SECRET domain and wild type CX3CL1 were mixed , left on ice for 1 h , and subjected to size exclusion column purification to obtain the SECRET/CX3CL1 complex . The SECRET domain and the SECRET/CX3CL1 complex were concentrated by ultrafiltration to ∼15 mg/ml . Crystals of the SECRET domain were grown from a mother liquor of 0 . 4 M Magnesium formate dehydrate , 0 . 1 M Bis-Tris propane , pH 7 . 0 with hanging-drop vapor diffusion method at room temperature . Crystals of the SECRET/CX3CL1 complex were grown from 0 . 2 M Lithium sulfate monohydrate , 0 . 1 M Tris·HC , pH 8 . 5 , 20% PEG4000 with hanging-drop vapor diffusion method at room temperature . Crystals of the SECRET domain were cryoprotected in well solution plus 20% ( v/v ) glycerol and cooled to 100 K before data collection . For the SAD data collection , crystals were soaked in well solution with 0 . 2 M NaBr for 30 s before data collection . Crystals of the SECRET/CX3CL1 complex were cryoprotected in well solution plus 20% ( v/v ) glycerol and cooled to 100 K before data collection . All diffraction data were collected at Shanghai Synchrotron Research Facility ( SSRF ) beamline BL17U . All data were indexed and integrated and scaled with program HKL2000 [52] . The structure of the SECRET domain was solved using the Br-SAD method . The positions of the Br were determined using the program SHELXD [53] and initial phases computed with the program SHELXE [54] as part of the HKL2MAP package [55] . Density modification was conducted using DM from the CCP4 suite [56] . The resulting electron density map was of excellent quality , allowing an automatic chain trance to be performed with the program Arp/wARP [57] . The following model adjustment and structural refinement were conducted using the program COOT [58] and PHENIX [59] , respectively . For the final model , the Rwork is 16 . 4% , and the Rfree is 19 . 9% . The structure of the SECRET/CX3CL1 complex was solved using the molecular replacement method with the SECRET domain and the CX3CL1 structures as search models in the program PHASER [60] . Iterative refinement with the program PHENIX [59] and model building with the program COOT [58] were conducted , yielding a final Rwork of 19 . 6% and Rfree of 25 . 0% . All structural figures were made by using PYMOL ( http://www . pymol . org ) . The binding affinity between the SECRET domain and CX3CL1 was determined by surface plasmon resonance ( SPR ) using BIAcore 3000 at 25°C . The SECRET domain was immobilized to about 350 Response Unit ( RU ) on a research-grade CM5 sensor chip in 10 mM sodium acetate , pH 4 . 1 by standard amine coupling method . The flow cell 1 was left blank as a reference . To measure binding affinity of CX3CL1 wild type and mutants by the SECRET domain , CX3CL1 in 10 mM HEPES , pH 7 . 2 , 150 mM NaCl , and 0 . 005% Tween-20 were injected over the flow cells at different concentrations at a flow rate of 30 µl min−1 . The binary complexes were allowed to associate for 90 s and dissociate for 90 s . The surfaces were regenerated with 5 mM NaOH between each cycle if needed . Data were analyzed with BIAcore 3000 evaluation software BIAevaluation 4 . 1 . To investigate the interference of heparin in the binding of CX3CL1 by the SECRET domain , 1 µM wild type CX3CL1 was pre-incubated with increasing concentrations ( 0 . 125 , 0 . 25 , 0 . 5 , 1 . 0 , 2 . 0 , 4 . 0 mg/ml ) of heparin sodium salt ( MW ∼15 , 000 Da , Sigma-Aldrich ) at 4°C for 1 h . SPR analysis was performed as above . To compare the binding ability of the SECRET domain wild type and mutant by chemokines , CX3CL1 purified by ourselves , CCL28 , CCL25 , and CXCL12 purchased from PeproTech were immobilized on the CM5 chip to ∼200 RU . SPR analysis was performed as above . RAW 264 . 7 cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated FBS at 37°C in CO2 incubator . Serum-starved RAW 264 . 7 cells with a total number of 1×107 were suspended in PBS buffer with 1 µM CellTracker Green CMFDA ( Invitrogen ) and incubated at 37°C for 5 minutes . The labeled cells were collected , washed three times with PBS buffer to remove the excessive CMFDA , and then suspended in RPMI 1640 medium for cell migration assays . Cell chemotaxis assay was performed using 8 µm−pore Cell Culture Inserts ( Millipore ) . The inserts were placed into 24-well plates containing RPMI1640 in the presence or the absence of CX3CL1 and SECRET domain . We seeded 8×104 CMFDA-labeled cells in each transwell insert and incubated at 37°C for 4 hours . Cell migration was quantified by counting the number of cells that remaining in the upper transwell by FACS . The coordinates of the SECRET domain and SECRET/CX3CL1 structures have been deposited into the Protein Data Bank with accession numbers 3ON9 and 3ONA , respectively . | Chemokines are a family of small proteins that help the immune system fight against invading pathogens by inducing the white blood cells to the areas of infection and inflammation . Due to the important roles of chemokines in immune response , the pathogens evolve diverse mechanisms to neutralize their activities . One example is that large DNA viruses , such as poxviruses and herpesviruses can produce chemokine binding proteins ( CKBPs ) to sequester chemokines during the infection . The SECRET domain represents a new family of viral CKBPs that was originally identified as a C-terminal extension of the viral tumor necrosis factor receptors ( vTNFRs ) . We determined the three-dimensional structures of the SECRET domain and its complex with chemokine CX3CL1 , revealing a new chemokine-binding manner of viral CKBPs . We also showed that other chemokines from different classes may be bound by the SECRET domain in a way similar to that observed in the SECRET/CX3CL1 complex structure . Our biochemical and chemotaxis assays also suggest that the SECRET domain is able to interfere with both chemokine-GAG and chemokine-receptor interactions , both of which are essential for chemokine activities in vivo . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biochemistry",
"viral",
"immune",
"evasion",
"proteins",
"virology",
"protein",
"structure",
"biology",
"microbiology"
] | 2011 | Structural Basis of Chemokine Sequestration by CrmD, a Poxvirus-Encoded Tumor Necrosis Factor Receptor |
The budding yeast spindle pole body ( SPB ) is anchored in the nuclear envelope so that it can simultaneously nucleate both nuclear and cytoplasmic microtubules . During SPB duplication , the newly formed SPB is inserted into the nuclear membrane . The mechanism of SPB insertion is poorly understood but likely involves the action of integral membrane proteins to mediate changes in the nuclear envelope itself , such as fusion of the inner and outer nuclear membranes . Analysis of the functional domains of the budding yeast SUN protein and SPB component Mps3 revealed that most regions are not essential for growth or SPB duplication under wild-type conditions . However , a novel dominant allele in the P-loop region , MPS3-G186K , displays defects in multiple steps in SPB duplication , including SPB insertion , indicating a previously unknown role for Mps3 in this step of SPB assembly . Characterization of the MPS3-G186K mutant by electron microscopy revealed severe over-proliferation of the inner nuclear membrane , which could be rescued by altering the characteristics of the nuclear envelope using both chemical and genetic methods . Lipid profiling revealed that cells lacking MPS3 contain abnormal amounts of certain types of polar and neutral lipids , and deletion or mutation of MPS3 can suppress growth defects associated with inhibition of sterol biosynthesis , suggesting that Mps3 directly affects lipid homeostasis . Therefore , we propose that Mps3 facilitates insertion of SPBs in the nuclear membrane by modulating nuclear envelope composition .
The hallmark feature of eukaryotic cells is the nucleus , a double membrane bound organelle that contains the genetic material . The outer nuclear membrane ( ONM ) of the nucleus is contiguous with the ER membrane while the inner nuclear membrane ( INM ) is distinct and contains a unique set of proteins that interact with chromatin and other nuclear factors . Embedded in the nuclear membrane are multiple nuclear pore complexes ( NPCs ) that regulate transport of macromolecules between the cytoplasm and the nucleus [1] . In organisms such as Saccharomyces cerevisiae that undergo a closed mitosis , the centrosome-equivalent organelle known as the spindle pole body ( SPB ) is present in the nuclear envelope throughout the life cycle [2] . The SPB organizes both cytoplasmic microtubules , which are involved in nuclear positioning , and nuclear microtubules , which are essential for chromosome segregation [3] . Both NPCs and SPBs are composed primarily of soluble proteins that partially assemble into sub-complexes in the nucleus or cytoplasm ( reviewed in [1] , [3] ) . Further assembly of both NPCs and SPBs requires insertion into the nuclear membrane at a point where the INM and ONM are joined together . Specific integral membrane proteins interact with soluble components of the NPC and SPB and are thought to anchor the complexes in the nuclear envelope . Ndc1 is essential for insertion of both the NPC and SPB [4]–[6] . At the NPC , three additional pore membrane proteins , Pom33 , Pom34 and Pom152 , play partially overlapping roles in NPC assembly [6]–[8] , while Nbp1 , Bbp1 and Mps2 are required in addition to Ndc1 for SPB insertion into the nuclear envelope [9]–[12] . The mechanism of NPC insertion has been extensively studied in both yeast and metazoan systems . Structural studies have shown that five subunits of the NPC ( Nup133 , Nup120 , Nup85 , Nup170 and Nup188 ) contain an ALPS motif ( for ArfGAP1 lipid packing sensor ) , which targets them to highly curved membranes [13] . These proteins are thought to form a coat complex on the nuclear envelope to facilitate NPC insertion [14]–[17] . In addition , membrane-bending proteins of the ER such as the reticulons have been shown to play a role in de novo NPC assembly [17] , [18] . Modification of lipids within nuclear membrane leaflets probably also occur at sites of NPC insertion to accommodate membrane curvature and fusion . Several proteins involved in lipid synthesis and membrane fluidity have been genetically linked to NPC assembly [19]–[21] , although their role in NPC insertion is not well characterized . In vertebrates cells , the SUN ( for Sad1-UNC-84 homology ) protein Sun1 also is required for NPC assembly [22] , [23] . A recent study suggested that hSun1 together with Pom121 is required for de novo assembly of NPCs possibly by facilitating membrane fusion [24] . The mechanism of SPB insertion into the nuclear membrane is poorly understood in comparison to insertion of NPCs . It is possible that many of the same events , such as membrane bending , curvature and lipid modification , are needed for SPB duplication since fusion of INM and ONM also must occur during SPB insertion . No specific factors that possess these functions have ever been directly implicated in the SPB duplication process with the exception of a recent report suggesting that the amphipathic alpha-helix of Nbp1 aids SPB insertion [12] . Perhaps one of the best clues as to how the SPB might insert into the nuclear membrane comes from a plethora of genetic interactions that have been identified between genes encoding SPB components and NPC subunits , including suppression of complete deletions of the SPB membrane components , MPS2 and MPS3 , by POM34 or POM152 deletion [4] , [25]–[27] . While it is possible that the NPC is involved in SPB insertion through its role in nuclear translocation of SPB subunits or in mRNA processing [25] , [28] , the fact that SPB duplication can occur in the absence of certain structural subunits points to a model in which NPCs and SPBs compete for a shared insertion factor , such as Ndc1 [4] . Alternatively , blocking NPC assembly by elimination of POM152 or POM34 could alter some aspect of the nuclear membrane that enables the SPB to duplicate in the absence of otherwise essential components . Consistent with this possibility , deletion of POM152 together with NUP170 results in nuclear membrane abnormalities [29] . In the present study , we further examine the role of the nuclear membrane in SPB duplication and show for the first time that changes in membrane composition are sufficient for SPB duplication in the absence of Mps3 . In addition , we demonstrate a role for the SUN protein Mps3 in regulation of membrane architecture and provide evidence that it functions in the insertion step of SPB duplication . This role of Mps3 is distinct from the previously described functions of Mps3 in SPB duplication . As a component of the SPB substructure that templates assembly for the new SPB , known as the half-bridge , Mps3 is required for initiation of SPB duplication and for tethering the half-bridge to the soluble core SPB through its interaction with the membrane protein Mps2 [30]–[32] . This function is similar to that of other SUN domain-containing proteins , which have been shown to play a role in accurate chromosome segregation and nuclear positioning due to their role in duplication and membrane tethering of centrosomes , basal bodies and SPBs in a wide variety of systems [33]–[35] . However , our current data suggest that SUN proteins have a novel function in membrane homeostasis , which is involved in insertion of protein complexes such as the SPB into the nuclear envelope .
A key feature of all SUN proteins , including Mps3 , is the presence of a number of structural motifs , including at least one transmembrane domain , regions of coiled-coils and a C-terminal SUN domain [35] . In addition , Mps3 contains an N-terminal acidic domain , a poly-glutamine region and a putative P-loop [30] ( Figure 1A ) . In order to determine their role in SUN protein localization and function , we created deletion alleles in each domain using site directed mutagenesis . In the case of the poly-glutamine repeat ( pQ ) , we not only deleted this region but also expanded it two-fold since this type of mutation is often linked to disease [36] , [37] . In the case of the putative P-loop , we mutated key residues anticipated to be involved in nucleotide binding to a residue with different chemical properties [38] , [39] . To test if the various mutants were functional , we assayed complementation of an mps3Δ mutant using a plasmid shuffle strategy . Although Mps3 plays multiple non-essential roles in the chromosome organization within the nucleus [28] , [40]–[49] , the primary function of Mps3 in mitotically dividing yeast cells is at the SPB [30] , [31] , so this test will allow us to identify domains of Mps3 that are essential for SPB duplication . mps3Δcc1 , mps3Δcc2 , mps3Δcc1Δcc2 and mps3ΔpQ mutants , which lack the first , second and both coiled-coil domains or the poly-glutamine region , respectively , did not exhibit any obvious growth defect at various temperatures ( Figure 1B; data not shown ) . To further examine their role in Mps3 function at the SPB , we tested for genetic interactions with mutants in other SPB components: mps2-1 , cdc31-2 , CDC31-16 , kar1Δ17 , spc42-11 , spc29-3 and sfi1-3 . Each of these mutants is either inviable ( synthetic lethal ) or shows an enhanced growth defect ( synthetic sick ) with mps3-1 , which has serine 472 in the SUN domain mutated to asparagine [30] . Like MPS3 , KAR1 and CDC31 encode components of the SPB half-bridge that are required for the initial step of SPB duplication: Kar1 is an integral membrane protein and Cdc31 is a small calcium binding protein that also binds to Sfi1 on the cytoplasmic side of the SPB half-bridge . Spc42 and Spc29 are components of the core SPB and Mps2 is a linker protein that tethers the half-bridge to the core SPB through interactions with Mps3 [3] , [32] . Surprisingly , only two genetic interactions were discovered with this new panel of mps3 alleles: synthetic lethality between mps3Δcc2 and kar1Δ17 , and suppression of the temperature sensitivity of the dominant CDC31-16 mutant by mps3ΔpQ ( Figure 1B ) . However , unlike mps3-1 mutants [30] , Cdc31 and Kar1 localization to the SPB was unaffected in mps3Δcc2 or mps3ΔpQ mutants ( data not shown ) . Most likely , this is because mps3Δcc1-GFP , mps3Δcc2-GFP and mps3ΔpQ-GFP localize to the SPB and INM in a pattern that is highly similar to that of Mps3-GFP ( Figure 1B ) . Taken together , our data suggests that mutation of the coiled-coil domains or deletion of the poly-glutamine region have at most minor effects on the localization and function of Mps3 during SPB duplication . Duplication of the poly-glutamine region did not affect cell growth ( Figure 1B ) . However , analysis of DNA content by flow cytometry revealed that mps3-2xpQ mutants exhibited an increase in ploidy ( Figure 1B ) , which is a common phenotype in SPB mutants , and it has been previously observed in a number of mps3 alleles [30] , [32] , [40] . The increase in ploidy was fully recessive as cells containing mps3-2xpQ and a wild-type copy of MPS3 were haploid ( data not shown ) , indicating that although the mps3-2xpQ protein may be only partially functional , it does not form a complex that titrates out other SPB duplication factors . Our observation that mps3-2xpQ-GFP levels at the SPB are reduced compared to Mps3-GFP ( Figure 1B ) suggests that this mutant is unable to localize to the SPB , perhaps due to a change in binding with its receptor at the SPB . Similar to mutants that eliminate or affect the SUN domain [30]–[32] , deletion of the transmembrane domain resulted in a non-functional version of Mps3 ( Figure 1B ) . This is most likely due to mislocalization of the mutant protein since mps3ΔTM-GFP was only visible in a diffuse pattern throughout the cytoplasm and the nucleus even in the presence of a wild-type untagged copy of Mps3 rather than at both SPBs and the peripheral nuclear envelope like Mps3-GFP ( Figure 1B ) [41] . Replacement of the Mps3 transmembrane domain with that of several other membrane proteins rescued the lethality of mps3Δ and restored localization to the SPB and the peripheral nuclear envelope ( Figure S1A and S1B ) . Interestingly , some of these chimeric proteins displayed significantly different localization patterns from wild-type Mps3-GFP ( Figure S1A ) . The fact that these proteins were sufficient to target enough Mps3 to the SPB to allow for cell proliferation and maintenance of genomic stability indicates that although membrane localization is critical for Mps3 function during SPB duplication , a specific transmembrane domain sequence is not required to target the SUN protein to the SPB . When point mutants were constructed in potential residues involved in nucleotide binding within the P-loop region , we found most alleles were able to complement mps3Δ and serve as the sole copy of MPS3 in the cell ( Figure 1B; data not shown ) , suggesting that this domain does not function in ATP-binding in vivo . However , some mutants in the P-loop region such as mps3-S190A displayed copy number sensitivity such that cells containing a single integrated copy of the mutant gene were viable , but cells containing two or more copies of the P-loop mutant gene were dead ( Figure 1D ) . This indicates that mps3-S190A is a weak dosage-sensitive antimorphic allele . We were unable to obtain transformants of one allele ( MPS3-G186K ) under a wide variety of conditions ( data not shown ) , suggesting that it is a dominant mutant that arrests cell growth . Using the galactose-regulatable GAL1-10 promoter , we set up a system so that we could examine the effects of MPS3-G186K on cell growth and SPB duplication . Cells containing a single integrated copy of GAL-MPS3 or GAL-MPS3-G186K in addition to the endogenous wild-type copy of MPS3 were analyzed for their effect on growth in a serial dilution assay at 30°C . Under these conditions , overexpression of wild-type MPS3 had a slight effect on cell growth while overexpression of MPS3-G186K significantly inhibited cell proliferation ( Figure 1E ) . This confirms that MPS3-G186K is a novel dominant lethal mutant . These same strains were examined following a 5 h induction with 2% galactose by flow cytometry to analyze DNA content and by indirect immunofluorescence microscopy with anti-alpha-tubulin and anti-gamma-tubulin antibodies to visualize microtubules and SPBs , respectively , to determine if the MPS3-G186K mutant affected SPB duplication and spindle assembly . We found that cells overproducing wild-type MPS3 did not arrest and underwent SPB duplication to form bipolar mitotic spindles ( Figure 2A , 2B ) . However , MPS3-G186K overproduction resulted in an accumulation of large-budded cells with a 2N DNA content , which is suggestive of a failure in SPB duplication and/or spindle formation ( Figure 2A ) . Examination of microtubule structures showed that 66% of MPS3-G186K mutants , but only 14% of GAL-MPS3 cells , contain monopolar spindles: a single DNA mass associated with a single SPB and microtubule array ( Figure 2B; n>200 ) . Unlike the monopolar spindle phenotype previously described for mps3 mutants where a single focus of gamma-tubulin is associated with the nuclear DNA [30] , [32] , 40% of MPS3-G186K cells contain two foci of gamma-tubulin , one of which does not nucleate microtubules ( Figure 2B ) . This phenotype is indicative of a defect in SPB insertion into the nuclear envelope: the SPB that does not nucleate microtubules has not properly inserted into the nuclear envelope and assembled inner plaque components , which are necessary for the formation of a nuclear microtubule array [3] . Although the uninserted SPB may nucleate cytoplasmic microtubules , these are generally difficult to observe due to the small number that is formed at each SPB . This phenotype is highly reminiscent of that observed in mutants that are defective in SPB insertion into the nuclear envelope such as mps2-1 , ndc1-1 , nbp1-dg and bbp1-1 [9]–[11] , [50] and suggests that like these SPB components , Mps3 has a function in the late step of SPB duplication in addition to its role in initiation of SPB duplication . In order for inner plaque components of the SPB such as Spc110 to assemble onto the newly duplicated SPB , the new SPB must be inserted into the nuclear envelope ( Figure 2C ) . Cells defective in SPB insertion due to a mutation in MPS2 or BBP1 contain two foci of the fluorescently-labeled central plaque component Spc42 , which is present at the old SPB and the duplication plaque/new SPB , but a single focus of fluorescently-labeled Spc110 [11] , [25] . In GAL-MPS3 cells , we found that in 80% of cells containing two Spc42-mCherry foci , those foci were coincident with two Spc110-GFP foci , indicating that these SPBs had duplicated and inserted into the nuclear envelope ( Figure 2D–2E ) . In contrast , only 61% of GAL-MPS3-G186K cells contained two SPBs that were labeled with both Spc42-mCherry and Spc110-GFP . 15% of the remaining cells contained a single SPB with both Spc42-mCherry and Spc110-GFP , indicative of an unduplicated SPB , 12% of cells contained two SPBs labeled with Spc42-mCherry only one of which co-labeled Spc110-GFP , indicative of an arrest at an intermediate step in SPB duplication , and 12% contained multiple SPB foci ( Figure 2D–2E ) . Thus , MPS3-G186K has pleiotropic effects on SPB duplication , including blocking SPB insertion into the nuclear envelope . Based on the fact that virtually all wild-type cells have two duplicated SPBs that co-labeled with Spc42-mCherry and Spc110-GFP , the increased frequency in aberrant pole morphologies that we observed in MPS3-G186K is striking and highly statistically significant ( p<0 . 01 ) . In addition , the levels of uninserted poles in MPS3-G186K is similar to that observed in other mutants such as mps2-1 and ndb1-dg , which fail in the late step of SPB duplication ( see [25] ) . We used electron microscopy ( EM ) to further evaluate the SPB duplication defect of cells overexpressing MPS3-G186K . Serial section analysis through the entire nucleus revealed that 11 out of 15 nuclei examined contained a single SPB . Of these 11 monopolar spindles , 4 had evidence of a “dead” pole characteristic of mutants defective in the late step of SPB insertion [9]–[11] , [50]; that is , an electron dense structure associated with the nucleus and with cytoplasmic microtubules but not inserted into the nuclear envelope ( Figure 2G ) . The remaining 7 SPBs appeared to be unduplicated , arresting with a terminal morphology similar to previously described mps3 alleles ( Figure 2F ) [30] , [32] . The remaining 4 cells contained short bipolar spindles ( Figure 2H ) . This ultrastructural analysis is consistent with our fluorescence and genetic data and suggests that MPS3-G186K affects multiple steps in SPB duplication , including SPB insertion into the nuclear envelope . We tested the ability of overexpressed SPB components to suppress toxicity of the mutant on galactose and found that 2μ-MPS3 partially rescued growth , confirming that MPS3-G186K is an antimorphic allele ( Figure S2 ) . Most other SPB components failed to rescue the growth arrest . The notable exception was 2μ-SPC42 . Spc42 is a component of the SPB central plaque that serves as a scaffold for assembly of the organelle and plays a role in anchorage in the nuclear membrane [51] , [52] . The fact that its overexpression partially restores growth to MPS3-G186K mutants is consistent with a defect in membrane insertion and tethering . In addition to the SPB duplication defect , several additional phenotypes were uncovered during the course of our EM analysis of MPS3-G186K cells that shed light onto the possible function of Mps3 in SPB insertion . As depicted in Figure 3C–3F , MPS3-G186K mutants appeared to have undergone massive over-proliferation of the nuclear membrane , resulting in 2–8 layers of nuclear envelope , and nuclei appear to have multiple lobes and extensions . The membrane expansion phenotype was highly specific and penetrant , occurring in 96% ( 48 of 50 ) of GAL-MPS3-G186K cells examined ( Figure 3C–3F ) but in none ( 0 of 34 ) of the GAL-MPS3 cells ( Figure 3A–3B ) . Therefore , it is most probably due to an effect of the MPS3-G186K mutant and not a general result due to overexpression of MPS3 or an integral membrane protein ( see also [31] ) . The fact that we observe excess membrane in MPS3-G186K mutant suggests that Mps3 directly or indirectly is involved in membrane homeostasis . Membrane proliferation in GAL-MPS3-G186K was restricted to the nucleus; no excess membrane was seen on other organelles , including the ER that is contiguous with the ONM in budding yeast ( Figure 4 ) . Previous work had suggested that the membrane adjacent to the nucleolus was subject to the formation of membrane flares [53] , but we found that all areas of the nuclear membrane underwent expansion , not just the nucleolar membrane . However , we did observe that the nucleolar region was often partitioned away from the main mass of the nucleus either by a membrane ( Figure 4B ) or by the formation of a lobe ( Figure 3C ) . Interestingly , within an individual nucleus , membrane proliferation was not uniform in that there were regions that contained a single bilayer and other regions containing multiple bilayers ( Figure 3C and 3D , Figure 4 ) . Analysis of serial nuclear sections showed that the excess membrane begins as tubules within the nucleus , which then proliferate underneath the existing membrane and then fuse with adjacent tubules ( Figure 4A ) or fold back upon itself to continue proliferation within the nucleus ( Figure 4B ) . The fact that the excess membrane forms only in tight association with existing nuclear envelope and does not form lamellae within the nucleus suggests that Mps3 is intimately involved in formation of the nuclear envelope layers . While it is possible that the excess membrane in MPS3-G186K inhibits SPB insertion , we found that SPBs as well as NPCs were often associated with a membrane region containing a single bilayer ( Figure 2F–2H and Figure 3C , 3E ) . However , we also detected NPCs in intermediate and inner layers , although these often appeared to be stacked with NPCs in adjacent layers ( Figure 3F ) , most likely to facilitate nuclear-cytoplasmic trafficking of macromolecules , which is reduced but not completely inhibited in MPS3-G186K ( Figure S3 ) . Unlike other mutants that affect the yeast nuclear membrane [18]–[21] , [54] , we did not observe partially assembled NPCs in the nucleoplasm or cytoplasm of MPS3-G816K mutants that would suggest a defect in NPC insertion , perhaps due to the large number of NPCs present in the nuclear envelope [55] and the relatively short period of time in which the dominant mutant is expressed ( Figure 3 and Figure 4 ) . Because both SPBs and NPCs are observed in nuclear membrane areas that contain a single bilayer and there is no apparent defect in NPC insertion , we suspect that the SPB assembly defect associated with MPS3-G186K is the result of membrane composition rather than the excess membrane . Membrane proliferation in MPS3-G186K mutants was also accompanied by an abnormal nuclear morphology . In many cases the nuclear membrane completely encircled regions of the cytoplasm , entrapping vesicles and other components ( Figure 4A–4B ) . Membrane extensions and protrusions were observed in all GAL-MPS3-G186K cells ( n = 50 ) and in 50% of GAL-MPS3 cells ( 17 of 34 ) , which do not over-proliferate the nuclear membrane ( Figure 3B–3D ) . The remaining 50% of GAL-MPS3 cells had a round or oval nuclear morphology similar to wild-type , depending on the cell cycle stage ( Figure 3A ) . We were able to follow the formation of nuclear morphology defects in real-time in GAL-MPS3-G186K mutants and GAL-MPS3 control cells using HDEL-dsRed to visualize the nuclear and ER membranes and Pus1-GFP to visualize the nucleus . Time-lapse image analysis showed that the nuclei in MPS3 overexpressing cells underwent minor changes in nuclear shape—virtually all nuclei were round or oval shaped throughout the 3 . 5 h time-course , with a few extensions and protrusions forming in some nuclei only at later time points ( Video S1; Figure 5B ) . In addition , we observed cells undergoing mitosis as the nucleus and nuclear membrane became elongated and then hour-glass shaped . In contrast , membrane extensions and nuclear deformation were easily observed in most MPS3-G186K cells by 1 . 5–2 h following addition of galactose ( Video S2 ) . At later time points , more severe membrane perturbation occurred in the mutant , often resulting in the formation of several masses of nuclear material within one cell ( Figure 5B ) . Although some mutant nuclei elongated , none completed mitosis to form two distinct masses of DNA . Thus , it appears that membrane proliferation is linked to an abnormal nuclear morphology and inhibition of mitotic progression . To better understand the role that Mps3 plays in SPB insertion and membrane structure , we screened the yeast deletion collection for mutants that rescued the growth defect of GAL-MPS3-G186K . If SPB duplication defects in this mutant are the result of defects in membrane composition as our cytology suggests , then it should be possible to find mutants that alter the lipid composition of the nuclear membrane to compensate for MPS3-G186K expression . In total , 93 mutants were found to grow on SC-URA+2% galactose in the presence of pURA3-GAL1-MPS3-G186K ( Table S2 ) . Because a number of these genes likely affect transcription , translation , post-translational modification or galactose-induction of the dominant allele , we used western blotting as a secondary screen to identify mutants that expressed high levels of Mps3-G186K protein , similar to that observed in wild-type cells ( data not shown ) . 37 mutants met this secondary criterion and are shown in Figure 6A together with a wild-type and gal4Δ control . GAL4 encodes a transcription factor that is required for activation of genes in response to galactose , including expression from GAL1 [56] . This mutant was identified in our primary screen , but as expected , failed to pass our secondary test for Mps3-G186K production and serves as a control ( Table S2 ) . Of the 37 deletion mutants that restore growth to MPS3-G186K , a number of genes encode proteins involved in transcription or translation . While these do not affect production of Mps3-G186K , they most likely restore growth to the mutant by affecting expression of an unknown target . One possible candidate target is POM34 , which has previously been shown to be regulated by translation and whose levels influence growth of mutants defective in SPB insertion , such as mps2 , bbp1 and ndc1 [4] , [25] , [26] . Interestingly , we also found that deletion of POM34 , POM152 and several other nucleoporins suppressed MPS3-G186K mutants ( Figure 6A ) . A number of these same deletion mutants were recently shown to rescue mps3Δ [26] . We proposed that deletion of the nucleoporins may rescue growth of cells lacking MPS3 by blocking NPC assembly , thereby liberating a shared insertion factor involved in both SPB and NPC insertion . Alternatively , their deletion may rescue growth of mps3Δ cells by changing the physical properties of the nuclear membrane , such as membrane fluidity , to facilitate SPB duplication without Mps3 . Our finding that MPS3-G186K mutants , which have defects in nuclear membrane structure and SPB duplication , are suppressed by the many of the same nucleoporin deletions as mps3Δ strongly suggests that changes in nuclear membrane properties , rather than liberation of an insertion factor , alleviate the growth defect in both cases . The fact that we isolated multiple nucleoporins as MPS3-G186K suppressors suggests a common mechanism of suppression . As we demonstrate below for pom152Δ and infer for the other nucleoporins , alteration of nuclear envelope properties appears to be responsible for suppression of MPS3-G186K as well as rescue of MPS3 deletion . In addition to nucleoporin deletions , we also found that deletion of two genes involved in lipid metabolism , FAA3 and DEP1 , rescues MPS3-G186K mutants . DEP1 encodes a transcriptional regulator of many metabolic genes , including genes involved in phospholipid biosynthesis [57] , [58] . FAA3 encodes one of five acyl coA synthetases that catalyze the conversion of fatty acids into activated acyl-coA intermediates in the first step in phospholipid biosynthesis . Faa1 , Faa2 , Faa3 , Faa4 and Fat1 localize to different subcellular compartments and display distinct specificities for medium , long and very-long chain fatty acids in vitro [59] . FAA3 encodes a long chain or very long chain fatty acyl coA synthetase that is believed to be partially redundant with Fat1 in vivo [60]–[62] . However , we found that deletion of the other acyl coA synthetases , including fat1Δ , was unable to suppress the growth defect of MPS3-G186K ( Figure 6B ) , suggesting that effects of Faa3 elimination are specific and that it may have a function in regulation of lipid composition at the nuclear membrane . Deletion or overexpression of FAA3 leads to an alteration in cellular lipid content [60] , [63] , suggesting that changes in fatty acid levels or composition are what rescue growth of MPS3-G186K cells . To test this idea , we treated cells with chemicals that alter membrane composition or fluidity and examined the effects on cell growth and nuclear envelope morphology . We found that growth of MPS3-G186K mutants on YPGR was rescued by addition of the membrane fluidizer benzyl alcohol , the sterol biosynthesis inhibitor terbinafine and the fatty acid oleic acid ( Figure 5A ) . In addition , increasing the temperature to 33°C , which increases membrane fluidity due to the cell's ability to adjust lipid composition by increasing levels of saturated and long chain fatty acids and altering the amount of sterols [64] , [65] , also suppressed the growth defect of MPS3-G186K , whereas decreasing the temperature to 23°C and 16°C , which decreases membrane fluidity , not only resulted in toxicity of MPS3-G186K but also exacerbated the phenotype associated with overexpression of wild-type MPS3 ( Figure 5A ) . Therefore , these data are consistent with the possibility that affecting lipid composition to increase membrane fluidity is sufficient to suppress MPS3-G186K . Our observation that overproduction of wild-type Mps3 is lethal at low temperatures suggests that GAL-MPS3 cells may have an altered membrane composition due to a direct role of Mps3 in membrane homeostasis . Further examination of cells by live cell imaging revealed that other defects associated with expression of MPS3-G186K , including abnormal nuclear morphology , cell division and SPB duplication , were at least partially suppressed by changing lipid composition . Oleic acid was particularly effective at suppressing defects in nuclear shape . The majority of nuclei appeared to maintain a round shape throughout a 3 . 5 h time course with few extensions or protrusions; however , few nuclei divided ( Figure 5B ) . A rare example is shown in Video S3 . In addition , analysis of DNA content by flow cytometry and examination of SPB insertion showed that most cells remained arrested in mitosis probably due to SPB duplication defects ( Figure 5C–5D ) . It is unclear why oleic acid suppresses growth of MPS3-G186K on plates but does not rescue in liquid culture , although similar effects have been reported for numerous mutants ( for example , [66] , [67] ) . Treatment of MPS3-G186K cells with other chemicals did not have as profound of an effect on nuclear morphology as oleic acid , perhaps because these drugs have pleiotropic effects on nuclear envelope structure ( Figure 5B , Videos S4 and S5 ) . For example , previous studies have shown that addition of benzyl alcohol affects NPC insertion [20] , [21] . While it may alleviate membrane expansion caused by MPS3-G186K , benzyl alcohol addition might result in defects in NPC structure that lead to changes in nuclear morphology . Nevertheless , despite terbinafine's partial rescue of nuclear shape , it eliminated the cell cycle arrest of MPS3-G186K and more than doubled the number of large-budded cells that contained two inserted SPBs ( Figure 5C–5D ) . Although this could be due to a slow growth phenotype associated with terbinafine since no mitotic divisions were observed ( Video S5 and Figure 5A ) , we favor the possibility that treatment with terbinafine affects sterol biosynthesis which in turn affects SPB duplication since we also observed that addition of other sterol inhibitors such as ketoconazole to media rescued growth of MPS3-G186K ( Figure S4 ) . Treatment with benzyl alcohol also only partially rescued the nuclear shape and cell cycle arrest of MPS3-G186K , but it was able to restore a distribution of SPB intermediates similar to that of GAL-MPS3 cells in galactose ( Figure 5B–5D ) , suggesting that it alleviates the SPB duplication defect associated with the dominant mutant . Consistent with this idea , we observed cells undergoing chromosome segregation and cell division in the presence of benzyl alcohol ( Video S4 ) . Importantly , addition of benzyl alcohol had virtually no effect on cell cycle progression or SPB insertion in GAL-MPS3 cells ( Figure 5C–D ) . We also found that addition of chemicals had similar effects on MPS3 and MPS3-G186K expression ( Figure S5 ) . Thus , the effects of MPS3-G186K on growth , nuclear morphology and SPB insertion can be rescued by altering the lipid composition of the cell using both genetic and chemical methods . Analysis of our mps3 deletion and point mutants showed that most are not sensitive to benzyl alcohol or sterol synthesis inhibitors ( data not shown ) , suggesting that the synthetic effects we observe as a result of these treatments on MPS3-G186K mutants are due to a specific defect in Mps3 function that may be related to SPB insertion since MPS3-G186K , but not other mutants , display a SPB duplication defect ( see Figure 1 and Figure 2 ) . In support of this hypothesis , we found that virtually all mutants in the C-terminal SUN domain , which is required for SPB duplication [30]–[32] , were resistant to terbinafine and found that a subset of these mutants , such as mps3-1 , mps3-A540D and mps3-W477A , were also sensitive to benzyl alcohol ( Figure 7A and 7B ) . Enhanced growth on terbinafine , which inhibits ergosterol biosynthesis [68] , suggests that cells containing the mps3-1 , mps3-A540D and mps3-W477A mutations have extremely fluid membranes; growth on membrane fluidizing agents like benzyl alcohol is toxic specifically to these mutants since they already have alterations in membrane composition . The more severe phenotype seen in some mps3 alleles presumably reflects a greater alteration in the nuclear membrane , although it is possible that some of these mutants may have other defects as well [32] , [40] . It is interesting that many , but not all , of the mps3 mutants that are sensitive to benzyl alcohol are alleles that spontaneously diploidize . Perhaps the larger size of the nucleus in these mutants requires a greater need for lipid synthesis that cannot be met with two mutant copies of MPS3 . To determine if changes in lipid composition generally affect the SPB duplication process or are specific to mps3 mutants , we compared the growth of different recessive mutants in SPB duplication on plates containing oleic acid , benzyl alcohol and terbinafine . Most SPB alleles tested , including mutants that disrupt early ( cdc31-2 , kar1-Δ17 , sfi1-3 ) , intermediate ( spc42-11 , spc29-3 ) and late steps ( mps2-1 , bbp1-1 , spc110-220 ) in SPB duplication and SPB regulators ( mps1-1 , tub4-1 ) [3] , grew at identical rates to wild-type under all conditions tested ( Figure 7B; data not shown ) . The notable exception was the ndc1-39 mutant , which is defective in both the late step of SPB duplication as well as in NPC assembly [5] . Like mps3-1 mutants , ndc1-39 cells exhibited increased sensitivity to oleic acid and benzyl alcohol and showed reduced sensitivity to terbinafine ( Figure 7B ) . This suggests that membranes in ndc1-39 , like mps3-1 mutants , have very fluid membranes due to changes in membrane composition . While it is possible that the basis for the membrane defect in ndc1-39 is related to its SPB insertion defect , we suspect that it is due instead to its role at the NPC since other SPB mutants , including mps2-1 and bbp1-1 , do not share this property . The fact that multiple mps3 mutants , including alleles that are defective in early steps of SPB duplication and not known to cause defects in NPC assembly or function [30] , [32] , are specifically affected by growth on chemicals affecting lipid composition strongly supports our hypothesis that Mps3 function is intimately linked to membrane homeostasis . Our finding that MPS3-G186K mutants exhibit a severe membrane over-proliferation phenotype that is rescued by drugs and mutants that affect lipid synthesis , together with our observations that certain mps3 mutants are sensitive to changes in membrane fluidity , indicates that Mps3 may play a role either directly or indirectly in regulating lipid composition in the cell . If this were the case , then we would anticipate that cells lacking MPS3 would have an altered lipid composition compared to wild-type cells . To test this hypothesis , levels of phospholipids and neutral lipids were examined in whole cell preparations from five biological replicates of wild-type , pom152Δ and pom152Δ mps3Δ cells grown to mid-log phase in YPD at 30°C by electrospray ionization tandem mass spectrometry ( ESI/MS/MS ) . pom152Δ mps3Δ mutants do not have an noticeable delay in cell division [26] , so analysis of lipid composition in these cells will not be complicated by changes in temperature , growth media or cell cycle arrest that are associated with other mps3 alleles . In cells lacking MPS3 , we found that there was no statistically significant change in the overall levels of polar lipids compared to the controls ( Figure 8A ) . However , examination of specific subclasses of phospholipids showed that pom152Δ mps3Δ mutants had 2 . 6 and 3 . 7 times the level of phosphatidyl serine ( PS ) and 2 . 0 and 2 . 3 times the level of phosphatidyl glycerol ( PG ) compared to wild-type and pom152Δ cells , respectively ( Figure 8A; Table S3 ) . Levels of phosphatidic acid ( PA ) and phosphatidyl ethanolamine ( PE ) were reduced 1 . 5 fold and increased 1 . 3 fold , respectively , in pom152Δ mps3Δ mutants compared with wild-type , although these values are not quite significant ( p = 0 . 11 for PA , p = 0 . 37 for PE ) . The levels of other phospholipids such as phosphatidyl choline ( PC ) and phosphatidyl inositol ( PI ) were largely unchanged in wild-type , pom152Δ and pom152Δ mps3Δ samples . Thus , the distribution of polar lipids is affected in cells lacking Mps3 function . Examination of neutral lipids revealed that cells lacking MPS3 contained 2–3 . 3 times the level of triacylglycerols ( TAGs ) compared to wild-type and pom152Δ cells ( Figure 8A; Table S3 ) . These cells also had an altered distribution of the three yeast sphingolipids , including a 1 . 7 fold decrease in inositol phosphorylceramide ( IPC ) and a 2 fold increase in mannosylinositol phosphorylceramide ( MIPC ) compared to controls , although these changes were not quite statistically significant ( p = 0 . 15 for IPC , p = 0 . 08 for MIPC ) . Even more importantly , we found that deletion of POM152 alone resulted in an 1 . 5 fold decrease in ergosterol levels as well as a reduction in diacylglycerol ( DAG ) , which was eliminated by simultaneous deletion of MPS3 ( Figure 8B; Table S3 ) . This finding is critical for several reasons . First , it provides biochemical support for our theory that nucleoporin deletions result in changes in the physical properties of the nuclear membrane ( [26] and above ) . Second , the fact that these changes in lipid composition are reversed by deletion of MPS3 is compelling evidence that Mps3 directly affects membrane homeostasis . From our lipidomics data , we would predict that Mps3 likely acts at multiple steps in lipid biosynthesis , affecting both neutral lipid and phospholipid levels . We would predict that changing properties of the nuclear membrane would result in growth defects in pom152Δ mps3Δ mutants if changes in lipid composition are required for SPB duplication and cell proliferation in the absence of MPS3 . Indeed , we found that pom152Δ mps3Δ mutants were sensitive to benzyl alcohol , particularly at 37°C ( Figure 9A ) . Although the lipid composition of pom152Δ mutants is altered , as was demonstrated in our lipidomics analysis ( Figure 8; Table S3 ) , growth of the single deletion is not affected by benzyl alcohol or other chemicals that affect membrane dynamics ( Figure 9A ) . However , overexpression of FAA3 , but not other acyl coA synthetases , partially rescued growth of pom152Δ mps3Δ mutants on benzyl alcohol-containing plates at high temperatures ( Figure 9B; data not shown ) . Therefore , it seems that Faa3 is involved in de novo lipid synthesis required for nuclear membrane homeostasis , in particular , in membrane changes brought about by Mps3 that are essential for SPB insertion .
Our analysis of the functional domains of Mps3 has led to new insights into its role in SPB duplication and nuclear envelope homeostasis . The requirement for the transmembrane domain in Mps3 function extends previous data suggesting that Mps3 is a structural component of the half-bridge [30] , [32] . However , our observation that multiple membrane domains can substitute for the Mps3 transmembrane domain is compelling evidence that no specific intra- or intermolecular interactions occur through this region of the protein , including transport through the NPC to the INM and anchorage at the SPB . We therefore tested the requirement for other conserved domains and were surprised to find that , at least when individually mutated , most were not essential for Mps3 function at the SPB in wild-type cells . Perhaps Mps3 has multiple binding partners and interaction domains due to the critical importance of SPB duplication in the maintenance of genomic integrity . Through mutagenesis of the putative P-loop region of Mps3 , we identified a new dominant allele that revealed a role for Mps3 in insertion of the newly duplicated SPB into the nuclear envelope . Although we could detect binding between Mps3 and ATP in vitro ( data not shown ) and glycine 186 is a conserved residue in the P-loop , we doubt that a defect in nucleotide binding is the underlying cause of the phenotypes we observed in MPS3-G186K mutants . This is because other mutants in conserved residues in the P-loop that also block ATP binding , such as S190A , G192K and S194A , did not result in defects in SPB duplication and nuclear envelope homeostasis . Also , alignment of Mps3 with other fungal orthologs shows that this motif is not conserved even within the Saccharomyces lineage . Thus , MPS3-G186K is likely a fortuitous antimorphic allele rather than a mutant that disrupts ATP binding per se . Although overexpression of MPS3 in our strains has few phenotypes at 30°C , its overproduction at 16°C and 23°C inhibits cell growth and leads to subtle changes in nuclear morphology ( see Video S1 ) . The simplest explanation for the temperature-dependent effect on cell viability when Mps3 is overproduced is that Mps3 plays a direct role in membrane homeostasis . Decreased amounts of saturated and long change fatty acids and changes in sterol levels , which accompany growth at lower at lower temperatures [64] , [65] , are incompatible with the alterations in membrane composition brought about by overproduction of Mps3 . This incompatibility leads to a failure in SPB insertion as well as other changes in nuclear structure . Changes in lipid composition and nuclear architecture are exacerbated in the MPS3-G186K mutant such that this allele is toxic under most conditions , including growth at 30°C . Moreover , because the membrane pertubations are striking and growth arrest is tight , this allele is amenable to genetic and cell biological analysis . Membrane over-proliferation in GAL-MPS3-G186K but not GAL-MPS3 strongly suggests that it is not simply due to overproduction of an integral membrane protein , but rather are a unique consequence of overexpressing a mutant version of MPS3 . We propose that this altered version of Mps3 titrates out key factors required for nuclear envelope homeostasis and/or SPB duplication . The most obvious candidate is the endogenous wild-type copy of Mps3 since SUN proteins are known to oligomerize and because a reduction of Mps3 at the spindle pole leads to defects in SPB duplication [23] , [32] , [69]–[71] . Consistent with this idea , mild overexpression of MPS3 using a 2-micron plasmid partially rescues the growth defect of GAL-MPS3-G186K in cells grown at 30°C ( Figure S2 ) . Other candidates include the ribosome biogenesis factors Erb2 and Rrs1 that interact with Mps3 and are required for nuclear morphology [48] . Proliferation of intracellular membranes in response to increased levels of membrane proteins has been previously reported in a wide range of cell types; however , the intranuclear membranes that we observed following expression of MPS3-G186K are structurally distinct in several ways . First , unlike the flattened arrays of ER membrane known as karmallae formed following overexpression of HMG1 , which encodes 3-hydroxy-3-methyl-glutaryl coenzyme A [72]–[75] , the MPS3-G186K-dependent membranes appear to be exclusively intranuclear and most likely represent overproliferation of the INM . Second , unlike the intranuclear membranes formed upon overproduction of Nup53 that are devoid of NPCs [54] , the MPS3-G186K membranes contain NPCs that are at least partially functional . In vertebrate cells , Sun1 localizes to NPCs and appears to play a role in their assembly [22]–[24] , however , this function does not appear to be conserved in yeast despite the fact that highly curved membranes are involved in both NPC and SPB duplication since Mps3 does not co-localize with NPCs [48] . Third , the intranuclear arrays that form following MPS3-G186K expression are generally closely associated with the nuclear membrane , suggesting that proteins on the surface of the MPS3-G186K-induced membranes are able to interact with other nuclear envelope proteins . This is in contrast to intranuclear membrane arrays that form within the nucleoplasm and do not appear to associate with the nuclear membrane over large regions [76] , [77] . Fourth , although the elaborate membrane extensions known as escapades formed by increasing levels of the peripheral membrane protein Esc1 also include NPCs [78] , the excess membrane found in escapades originates at the nuclear vacuolar junction and results in predominant cytoplasmic extensions , whereas the MPS3-G186K membranes appear to begin as tubules inside of the nucleus that then fuse into intranuclear membranes . Fifth , unlike the membrane flares formed by disruption of proteins involved in the production of PA and DAG [53] , [79]–[81] , membrane proliferation in MPS3-G186K mutants was not restricted to the nucleolar region but rather appeared to occur at multiple sites inside of the nucleus . Interestingly though , the nucleolus was often partitioned away from the nucleus in a separate lobe in cells overexpressing both MPS3-G186K as well as MPS3 , consistent with the idea that this region of the nuclear membrane has a unique molecular composition . In cells overexpressing MPS3 or MPS3-G186K , we observed increased levels of most types of polar lipids as well as ergosterols compared to wild-type ( data not shown ) . Levels of DAG were also elevated in MPS3 overexpressing cells , but not in MPS3-G186K mutants , which might account for some of the differences we observed in membrane morphology and SPB duplication between these two strains . In MPS3-G186K cells , the excess membrane appears to form as tubules within the nucleoplasm that then fuse to form the intranuclear membranes observed in the mutants . It is unclear how these tubules initiate , but they appear to be intermediates in nuclear envelope assembly . Curiously , similar tube-like structures have been detected in cells overexpressing NUP53 and during the formation of nuclear membranes around sperm chromatin in Xenopus oocytes [54] , [82] . Our observation that lipid composition affects cell viability , nuclear shape and membrane morphology of MPS3-G186K mutants suggests that specific properties of intranuclear tubules are critical for their formation , fusion and stacking to form an intact nuclear envelope . At least one important parameter is the amount of fatty acids produced by Faa3 . This enzyme catalyzes acylation of long and very long chain fatty acids in vitro , which are important for the formation of highly curved membranes such as those that occur at sites of NPC and SPB insertion [83] , [84] . By limiting the amount of these classes of fatty acids , it may be more difficult to form a membrane bend , which could help limit the membrane extension and protrusion driven by MPS3-G186K . We did not detect a difference in chain length in fatty acids in any of our mps3 mutants , however ( Table S3; data not shown ) . Possibly only a small pool of fatty acid side chains is modified making detection in a population assay difficult , or Faa3 may have a different substrate specificity in vivo . In addition , other membrane modifications such as sterol insertion may be a driving force in the control of nuclear envelope structure inside the cell . Many of the genes involved in sterol biosynthesis in yeast are encoded by essential genes [85] and thus , they are not present in the haploid yeast deletion collection so they would not have been discovered in our screen . Deletion of NPC subunits suppresses the SPB duplication defect associated with mutation of several genes encoding SPB components [4] , [25]–[27] . Our finding that deletion of POM152 results in reduced levels of ergosterols and DAG compared to wild-type cells points to the possibility that the nucleoporin deletions affect the lipid composition of the nuclear envelope . This might occur through reduced transport of genes involved in lipid biosynthesis or through decreased export of the associated mRNAs . These changes in membrane composition could also account for genetic interactions between pom152Δ and genes involved in membrane fluidity , sterol synthesis , organelle integrity and membrane bending [18] , [20] , [21] , [86] . In pom152Δ mps3Δ mutants , lipidomics analysis shows that the balance of ergosterol and DAG is reset to near wild-type levels , which could explain why SPB duplication and nuclear morphology defects were not observed in the double delete [26] . Other classes of lipids such as TAG are elevated in cells lacking MPS3 , suggesting that Mps3 plays a direct role in regulation of membrane homeostasis . Our observation of allele specific defects in membrane proliferation and drug sensitivity in mps3 mutants that was not detected in other SPB mutants further supports this hypothesis . The fact that these same alleles of MPS3 are defective in SPB duplication indicates that Mps3-dependent changes in the nuclear envelope are important for formation of the new SPB , perhaps at multiple steps in SPB assembly . Studies in yeast , Dictyostelium , C . elegans and mammalian cells have shown that loss of SUN protein function leads to aberrant nuclear morphology [30] , [32] , [41] , [48] , [87]–[89] . Whether they are directly involved in the synthesis or maintenance of the nuclear membrane or are indirectly involved through binding to the nuclear lamina is not well understood [90] . Perhaps the best hint as to how SUN proteins might function in membrane homeostasis comes from two-hybrid analysis of the fission yeast SUN protein Sad1 , which pointed to an interaction with the acetyl coA carboxylase enzyme Cut6 that is needed for de novo biosynthesis of long-chain fatty acids and mitotic progression [91] , [92] . Much like the binding between SUN proteins and their ONM binding partners , known as KASH proteins , in the perinuclear space [35] , we envision that SUN proteins like Mps3 may also associate with factors involved in membrane structure . Mps3 binding to enzymes or other proteins involved in membrane remodeling could be a mechanism to promote membrane bending and fusion at specific sites in the nuclear envelope , such as at the duplicating SPB or the NPC . However , Mps3 could also control membrane synthesis at the transcriptional level , by sequestering factors involved in lipid synthesis at the nuclear periphery . Future studies aimed at identification of Mps3 binding partners will help us better understand the role SUN proteins in membrane dynamics and nuclear morphology as well as elucidate the mechanism by which Mps3-dependent changes in the nuclear membrane enable duplicated SPBs to insert into the nuclear envelope .
All strains are derivatives of W303 ( ade2-1 trp1-1 leu2-3 , 112 ura3-1 his3-11 , 15 can1-100 ) and are listed in Table S1 except those used in Figure 6 and Table S2 , which are in BY4741 ( Mata his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 ) and were taken from the yeast deletion collection ( Open Biosystems ) . Standard techniques were used for DNA and yeast manipulations . Deletion of MPS3 , fusion of SPC110 and NUP49 to GFP and fusion of HTB2 ( which encodes one of the two copies of histone 2B ) , SPC42 or NET1 to mCherry was done by PCR-based methods [93] , [94] . Correct integration was confirmed by PCR . The MPS3 ORF and ∼500 bp of promoter were amplified by PCR and cloned into the XhoI and BamHI sites of a pRS306 vector containing GFP to construct pSJ650 ( pRS306-MPS3-GFP ) [41] . Construction of pSJ148 ( pRS305-MPS3 ) has been previously described [32] . To overexpress MPS3 , the entire open reading frame was inserted immediately adjacent to GAL1 in pRS306 to create pSJ146 ( pRS306-GAL1-MPS3 ) . Plasmids were digested with BstEII or ApaI to target integration to LEU2 or URA3 , respectively . The number of copies of MPS3 integrated was determined by Southern blotting . Deletion and point mutants , as well as transmembrane insertions , were generated in these plasmids using the Quick Change Mutagenesis Kit ( Stratagene ) . Each deletion mutant contains an in-frame deletion of the indicated amino acids; in the case of 2xpQ mutant , an extra copy of the pQ domain was inserted immediately after the pQ coding sequence . Sequencing was performed to confirm correct base pair substitutions or deletions were made . To test for rescue of MPS3-G186K , plasmids containing the indicated SPB genes were taken from the yeast tiling library [95] and transformed into SLJ1797 . For dilution assays , 5 OD600 of cells were serially-diluted 10-fold in sterile growth media and stamped onto agar plates . YPD has 2% glucose and YPGR has 2% galactose and 2% raffinose as the carbon source . Chemicals were purchased from Sigma and were added to media in the following amounts: 5 mM oleic acid , 0 . 2% benzyl alcohol , 1 . 25 µg/ml terbinafine and 1 µg/ml ketoconazole . Localization of Mps3-GFP , H2B-mCherry , the nuclear localization sequence ( NLS ) reporter , HDEL-dsRed and Pus1-GFP were visualized as previously described [41] . Briefly , 1 ml of culture was centrifuged , washed with 1 ml ddH2O , resuspended in approximately 100 µL of ddH2O and 10 µL was placed on a 25% gelatin pad for image analysis . Fluorescence was performed using a confocal microscope ( LSM-510-META; Zeiss ) equipped with a ConforCor 3 module with avalanche photodiode detectors , which allow single photon counting , with a 100× 1 . 46 NA α-Plan Fluar objective ( Zeiss ) . GFP was excited using a 488-nm Argon laser line , while mCherry was excited with a 561-nm HeNe laser line with the appropriate filter sets . Emitted photons were collected through BP 505–540 nm and LP 580 nm filters , with a pinhole size of 1 . 03 Airy units according to the green channel . Data was acquired using AIM v . 4 . 2 software ( Zeiss , Inc . ) . Images were collected with 8–10 image stacks with a 0 . 3 micron step size through the cells at room temperature . Images were processed using ImageJ software ( NIH ) . At least two independent transformants of each genotype were analyzed by fluorescence microscopy in at least three independent experiments . N∶C ratios were calculated as previously described [28] . Time-lapse image analysis of GAL-MPS3 or GAL-MPS3-G186K strains was performed on cultures grown overnight in YEP plus 2% raffinose at 30°C then diluted back to an OD600 of 0 . 2 and grown in the same media for 2 h . 2% galactose and nothing , 0 . 2% benzyl alcohol , 5 mM oleic acid , 1 . 25 µg/ml terbinafine or 1 µg/ml ketoconazole were added at 30°C for 30 min . One ml of culture was centrifuged to concentrate the cells and 5 µL was placed on a 1% agarose pad containing the same media and chemical treatment . Fluorescent images were taken from 1 h to 3 h 30 min after galactose induction using the same microscope configuration but with a 40× 1 . 3 NA oil objective ( Zeiss ) . Emitted photons were collected through BP 505–540 nm and LP 580 nm filters , with a pinhole size of 1 . 03 Airy units according to the green channel . Images were collected with 17 image stacks with a 0 . 7 micron step size through the cells at room temperature . The 512×512 images were binned spatially 2×2 and a max projection was applied for each Z-slice . All data processing and video generation were done in ImageJ ( NIH ) . For the characterization of the SPB duplication using the red/green foci assay , images were acquired with a 100× 1 . 4 NA oil objective on an inverted Zeiss 200 m equipped with a Yokagawa CSU-10 spinning disc . 488 nm excitation and 568 nm excitation were used for GFP and mCherry , respectively , and emission was collected through BP 500–550 nm and BP 590–650 nm filters , respectively , onto a Hamamatsu EMCCD ( C9000-13 ) . For each channel , a Z-stack was acquired using 0 . 6 or 0 . 7 micron spacing . 13 total slices were acquired , and a max projection image was created for analysis of foci using ImageJ ( NIH ) . Analysis of DNA content by flow cytometry , EM , and protein localization by indirect immunofluorescence microscopy were performed as previously described [30] , [32] , [96] . Poly ( A ) + in situ hybridization was performed using an Cy3-labeled oligo ( dT ) 50 probe as described [97] . Cells were examined with a Zeiss Axioimager using a 100× Zeiss Plan-Fluar lens ( NA = 1 . 45 ) , and images were captured with a Hamamatsu Orca-ER digital camera and processed using ImageJ ( NIH ) . The following primary antibody dilutions were used: 1∶1000 anti-Flag ( Sigma ) and 1∶1000 anti-glucose-6-phosphate dehydrogenase ( G6PDH; Sigma ) . Fluorescently conjugated secondary antibodies were used at 1∶10000 for analysis and quantification using the Odyssey system ( Li-Cor ) . Yeast cells were grown overnight at 30°C in YPD , diluted to OD600 of 0 . 2 in the same media and allowed to grow for 6–8 h at 30°C . The wet weight of the sample was recorded and lipids were extracted by the following procedure . Pelleted cells were homogenized using glass beads and water and the aqueous homogenized cells were transferred to 15 ml Teflon-lined glass tubes . To 0 . 8 parts ( 1 . 6 ml ) of homogenized cells in aqueous solution , 1 part ( 2 ml ) of chloroform and 2 parts ( 4 ml ) of methanol were added . The slurry was shaken , 1 part ( 2 ml ) of chloroform and 1 part of water ( 2 ml ) were added . After shaking well , the slurry was centrifuged at low speed for 5–10 min and the lower layer was removed . One part of chloroform was added , shaken , centrifuged and the lower layer was again removed . This was repeated , and the combined lower layers were washed once with a small volume of 1 M KCl and once with a small volume of water . The combined and washed chloroform extracts were dried under liquid nitrogen and stored at −20°C in 2 ml Teflon-lined glass vials . An automated electrospray ionization-tandem mass spectrometry ( ESI-MS/MS ) approach was used to analyze lipid composition in these samples , and data acquisition and analysis were carried out as described previously [98] with modifications . A 5 to 20 µl of aliquot of extract in chloroform was used from each 1 ml lipid sample . Precise amounts of internal standards , obtained and quantified as previously described [99] , were added in the following quantities: 0 . 3 nmol di12:0-PC , 0 . 3 nmol di24:1-PC , 0 . 3 nmol 13:0-lysoPC , 0 . 3 nmol 19:0-lysoPC , 0 . 3 nmol di12:0-PE , 0 . 3 nmol di23:0-PE , 0 . 3 nmol 14:0-lysoPE , 0 . 3 nmol 18:0-lysoPE , 0 . 3 nmol di14:0-PG , 0 . 3 nmol di20:0 ( phytanoyl ) -PG , 0 . 3 nmol di14:0-PA , 0 . 3 nmol di20:0 ( phytanoyl ) -PA , 0 . 2 nmol di14:0-PS , 0 . 2 nmol di20:0 ( phytanoyl ) -PS , 0 . 46 nmol 16:0-18:0-PI , 0 . 33 nmol di18:0-PI , 3 . 1 nmol tri17:1 TAG , and 4 . 6 nmol di15:0 DAG . The sample and internal standard mixture was combined with solvents , such that the ratio of chloroform/methanol/300 mM ammonium acetate in water was 300/665/35 , and the final volume was 1 . 4 ml . Unfractionated lipid extracts were introduced by continuous infusion into the ESI source on a triple quadrupole MS ( 4000QTrap , Applied Biosystems ) . Samples were introduced using an autosampler ( LC Mini PAL , CTC Analytics AG , Zwingen , Switzerland ) fitted with the required injection loop for the acquisition time and presented to the ESI needle at 30 µl/min . Sequential precursor and neutral loss scans of the extracts produce a series of spectra with each spectrum revealing a set of lipid species containing a common head group fragment . Lipid species were detected with the following scans: PC and lysoPC , [M+H]+ ions in positive ion mode with Precursor of 184 . 1 ( Pre 184 . 1 ) ; PE and lysoPE , [M+H]+ ions in positive ion mode with Neutral Loss ( NL ) of 141 . 0 ( NL 141 . 0 ) ; PG , [M+NH4]+ in positive ion mode with NL 189 . 0 for PG; PI , [M+NH4]+ in positive ion mode with NL 277 . 0; PS , [M+H]+ in positive ion mode with NL 185 . 0; PA , [M+NH4]+ in positive ion mode with NL 115 . 0; IPC , [M−H]− in negative ion mode with Precursor of 259 . 0; MIPC , [M−H]− in negative ion mode with Precursor of 421 . 0; M ( IP ) 2C , [M−H]− in negative ion mode with Precursor of 663 . 1; ( lysoPG , [M−H]− in negative ion mode with Precursor of 152 . 9 used as a standard for IPC , MIPC , M ( IP ) 2C ) . The collision gas pressure was set at 2 ( arbitrary units ) . The collision energies , with nitrogen in the collision cell , were +28 V for PE , +40 V for PC , +25 V for PA , PI and PS , +20 V for PG , −72 for IPC , −80 for MIPC , −75 for M ( IP ) 2C , −57 for lysoPG . Declustering potentials were +100 V for all positive mode scans , −180 for IPC , MIPC , and M ( IP ) 2C , −100 for lysoPG . Entrance potentials were +15 V for PE and +14 V for PC , PA , PG , PI , and PS , −13 for IPC , MIPC , and M ( IP ) 2C , −10 for lysoPG . Exit potentials were +11 V for PE and +14 V for PC , PA , PG , PI , PS , −10 for IPC , MIPC , and M ( IP ) 2C , −14 for lysoPG . The scan speed was 50 or 100 u per sec . The mass analyzers were adjusted to a resolution of 0 . 7 u full width at half height . For each spectrum , 9 to 150 continuum scans were averaged in multiple channel analyzer ( MCA ) mode . A series of neutral loss scans were used to detect ergosterol esters , DAG , and TAG species as [M+NH4]+ ions . The scans targeted losses of various fatty acids as neutral ammoniated fragments: NL 285 . 2 ( 17∶1 , for the TAG internal standard ) ; NL 259 . 2 ( 15∶0 , for the DAG internal standard ) ; NL 271 . 2 ( 16∶1 ) ; NL 273 . 2 ( 16∶0 ) ; NL 299 . 2 ( 18∶1 ) ; and NL 301 . 2 ( 18∶0 ) . The scan speed was 100 u per sec . The collision energy was +25 V , entrance potential was +14 V , and exit potential was +14 V . Fifty continuum scans were averaged in MCA mode . The source temperature ( heated nebulizer ) was 100°C , the interface heater was on , +5 . 5 kV was applied to the electrospray capillary , the curtain gas was set at 20 ( arbitrary units ) , and the two ion source gases were set at 45 ( arbitrary units ) . The background of each spectrum was subtracted , the data were smoothed , and peak areas integrated using a custom script and Applied Biosystems Analyst software , and the data were corrected for overlap of isotopic variants ( A+2 peaks ) . The first and typically every 11th set of mass spectra were acquired on the internal standard mixture only . Peaks corresponding to the target lipids in these spectra were identified and molar amounts calculated in comparison to the two internal standards on the same lipid class . To correct for chemical or instrumental noise in the samples , the molar amount of each lipid metabolite detected in the “internal standards only” spectra was subtracted from the molar amount of each metabolite calculated in each set of sample spectra . The data from each “internal standards only” set of spectra was used to correct the data from the following 10 samples . Finally , the data were corrected for the fraction of the sample analyzed and normalized to the mg wet weight to produce data in the units nmol/mg , except DAG , TAG , and ergosterol esters . For analysis of these classes , there is variation in ionization efficiency [100] . Thus , the ratio of the MS response of the internal standard does not provide a value that is directly proportional to the content of each species . DAG , TAG , and ergosterol ester amounts are thus expressed as relative mass spectral signal/mg wet weight , where a signal of 1 . 0 represents a signal equal to the signal of 1 nmol tri17:1 TAG , or 1 nmol di15:0 DAG ( the internal standards ) . | Accurate segregation of chromosomes during mitosis is essential to prevent genetic instability and aneuploidy that lead to cancer and other diseases . Centrosomes and spindle pole bodies mediate the assembly of a microtubule-based structure known as the mitotic spindle , which physically separates chromosomes during mitosis so that the two daughter cells contain a complete copy of the genetic material as well as a spindle pole . During every cell cycle , the DNA and the spindle pole must be duplicated exactly once to ensure proper formation of a bipolar mitotic spindle . In yeast cells , the nuclear envelope does not break down , so the spindle pole must be inserted into the nuclear membrane so that it can form both the microtubules involved in the mitotic spindle and those involved in positioning of the nucleus . How a large protein complex such as the spindle pole body is inserted into the lipid layers of the nuclear membrane is not well understood . We show that the evolutionarily conserved SUN protein Mps3 is involved in spindle pole insertion into the nuclear membrane . This likely reflects a function for SUN proteins in controlling nuclear envelope structure by modulating the types of lipids that are present in the nuclear membrane . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microbiology",
"chemical",
"biology",
"mitosis",
"histology",
"model",
"organisms",
"cell",
"growth",
"molecular",
"genetics",
"chromosome",
"biology",
"membranes",
"and",
"sorting",
"chemistry",
"biology",
"cell",
"biology",
"genetic",
"screens",
"genetics",
"yeast",
... | 2011 | The SUN Protein Mps3 Is Required for Spindle Pole Body Insertion into the Nuclear Membrane and Nuclear Envelope Homeostasis |
In insects , products of the male reproductive tract are essential for initiating and maintaining the female post-mating response ( PMR ) . The PMR includes changes in egg laying , receptivity to courting males , and sperm storage . In Drosophila , previous studies have determined that the main cells of the male accessory gland produce some of the products required for these processes . However , nothing was known about the contribution of the gland's other secretory cell type , the secondary cells . In the course of investigating the late functions of the homeotic gene , Abdominal-B ( Abd-B ) , we discovered that Abd-B is specifically expressed in the secondary cells of the Drosophila male accessory gland . Using an Abd-B BAC reporter coupled with a collection of genetic deletions , we discovered an enhancer from the iab-6 regulatory domain that is responsible for Abd-B expression in these cells and that apparently works independently from the segmentally regulated chromatin domains of the bithorax complex . Removal of this enhancer results in visible morphological defects in the secondary cells . We determined that mates of iab-6 mutant males show defects in long-term egg laying and suppression of receptivity , and that products of the secondary cells are influential during sperm competition . Many of these phenotypes seem to be caused by a defect in the storage and gradual release of sex peptide in female mates of iab-6 mutant males . We also found that Abd-B expression in the secondary cells contributes to glycosylation of at least three accessory gland proteins: ovulin ( Acp26Aa ) , CG1656 , and CG1652 . Our results demonstrate that long-term post-mating changes observed in mated females are not solely induced by main cell secretions , as previously believed , but that secondary cells also play an important role in male fertility by extending the female PMR . Overall , these discoveries provide new insights into how these two cell types cooperate to produce and maintain a robust female PMR .
The homeotic transcription factor Abdominal-B ( Abd-B ) specifies the identity of the four most-posterior abdominal segments of the fly ( the 5th through 8th abdominal segments ) , as well as the genital and anal structures [1]–[3] . Each of these segments is specified by a particular pattern and level of Abd-B protein expression in the early embryo . Four segment-specific cis-regulatory domains ( iab-5 through iab-8 ) spanning >90 kb of DNA have been shown to control the expression pattern of Abd-B , where each domain is predominantly responsible for controlling the Abd-B expression pattern in one particular segment [4]–[6] ( for a review see [7] ) . Extensive study has been devoted to exploring how the segment-specific expression pattern of Abd-B is achieved . Due to the striking cuticular transformations elicited by Abd-B mutations , genetic and transgenic analyses have been able to discover numerous enhancers , silencers and insulators that direct Abd-B expression in the ectoderm [8]–[16] . However , much less is known about the role of Abd-B in non-ectodermally derived tissues during later stages of development . Here , we use a 111 kb BAC-based reporter construct to identify new locations of Abd-B expression in the adult fly . We find that Abd-B is strongly expressed in the accessory gland ( AG ) , a secretory tissue of the adult male reproductive tract that has important reproductive functions . The AG synthesizes seminal proteins that are essential for male fertility . These >180 accessory gland proteins ( “Acps” ) are transferred to females during mating and cause post-mating changes in the females known collectively as the post-mating response ( PMR ) . The PMR includes increased rates of egg-laying and ovulation , sperm storage , decreased receptivity to courting males , as well as changes in longevity , feeding , and sleep patterns ( reviewed in [17] , [18] ) . The PMR is divided into two phases . The short term response ( STR ) refers to changes in the above behaviors during the first ∼24 hours post-mating . It requires Acps , but not the receipt of sperm . Persistence of the PMR after 24 hr ( and for up to ∼10 days ) is known as the long-term response ( LTR ) . The LTR requires Acps and stored sperm [19]–[22] . Many of the roles of Acps were initially discovered by experiments in which whole AG extracts or purified Acps were injected into unmated females [23]–[25] , or by whole-tissue ablation in males [26] . Each lobe of the AG is composed of a monolayer of approximately 1000 secretory cells comprised of two morphologically distinct cell types . Roughly 96% of these cells are flat , polygonally shaped “main cells” . The remaining 4% of the cells are large , spherical , vacuole filled “secondary cells ”; these are dispersed among the main cells at the distal tip of the gland . Enhancer trapping and other studies have shown that , in addition to their morphological differences , these two secretory cell types are biochemically distinct [27]–[29] . Ablation of the main cells only [19] showed that products of these cells are essential for the PMR . These products include ovulin ( Acp26Aa ) , an Acp that acts in the STR to stimulate ovulation [30] , [31] , and the sex peptide ( SP , Acp70A ) , which is the ultimate regulator of most other PMR effects [22] , [32]–[35] . SP binds to sperm within the mated female , and its active portion is gradually released from the sperm [22] . This binding and release allows SP to affect the female for as long as she contains stored sperm . A network of five other Acps is necessary for SP to bind to sperm and enter storage . The predicted protease CG10586 ( Seminase ) [36] appears to be necessary for both STR and LTR related events , while the predicted protease CG9997 , the predicted cysteine-rich secretory protein ( CRISP ) CG17575 , and the predicted lectins CG1656/1652 appear to be LTR specific [37]–[40] . The cellular source of each of these proteins is currently unknown . In spite of the detailed characterization of the main cells and several specific Acps , the role of the secondary cells has remained mysterious . No PMR-associated Acps were known to be expressed exclusively in the secondary cells , and no tools have been available to specifically target those cells . Here , we identified the secondary cells of the male AG as a novel location of Abd-B expression in the adult fly . By screening an extensive collection of cis-regulatory deletions [6] , [41] , [42] , we discovered a 2 . 8 kb enhancer from the iab-6 cis-regulatory domain , whose removal completely abolishes Abd-B expression in the secondary cells . Loss of Abd-B expression in the secondary cells causes those cells to develop aberrantly . Moreover , these mutant males provide their mates with substances that initiate the PMR , but are insufficient to maintain it . Our results indicate that Abd-B expression in the secondary cells is essential for their proper development and for the production of proteins important for long-term changes in female post-mating responses .
In order to discover new tissues in which the Abd-B gene functions , we undertook the creation of a transgenic reporter that accurately reproduces the Abd-B expression pattern throughout development . Previous studies indicated that the Abd-B gene is expressed as two isoforms , the Abd-B m and r forms , and that the expression of these two isoforms requires separate elements located within a large cis-regulatory region spanning >90 kb of DNA [43] . As the Abd-Br isoform is thought to be primarily involved in the formation of the external genitalia [44] , we decided to concentrate our study on the Abd-Bm isoform , which is involved in determining segment identity . BACR24L18 is a BAC of ∼172 kb that contains the Abd-B , and much of the abd-A region of the Bithorax complex ( BX-C ) . By recombineering , we reduced BAC24L18 to contain mostly the iab-5 to iab-8 domains required for Abd-Bm expression ( removing many of the Abd-Br alternative promoters and its regulatory elements ) and the Abd-Bm coding sequence ( Figure 1B ) . A ΦC31 AttB integration sequence and a white integration marker were also added during the reduction step ( Figure 1B and 1C ) . We first tested if expression derived from the sequences on this BAC were sufficient to rescue Abd-B mutant phenotypes . We integrated the Abd-B BAC into the 51C landing platform [45] and tested for complementation of two large deletions affecting Abd-B activity . We found that the presence of a copy of the BAC on the second chromosome rescues the mutant phenotypes of iab-6 , 7IH and iab-5 , 6J82 [6] ( Figure S1; See Text S1 ) . Because the sequences preserved on the BAC seemed to drive appropriate Abd-Bm expression , we proceeded to modify the BAC by recombineering to replace the first codon of the first exon of Abd-B with the sequence encoding the Gal4 transcription factor . As this sequence also adds a stop codon , the expression of Abd-B from the BAC should be eliminated , but any sequences that might be used in Abd-B gene regulation will be preserved to drive reporter gene expression . The final BAC used in the experiments was 111 kb ( Figure 1 ) . It was integrated into the 51C landing platform . To study the Abd-B expression pattern , a line was established containing the Abd-B-Gal4 BAC and a UAS-GFP reporter . Initial examination of the embryonic expression pattern in these lines confirms that the Abd-B-Gal4 BAC appears to recapitulate most of the wild-type expression pattern of Abd-Bm in early embryos ( Figure 2A; Figure S2; Figure S3 ) . Later , we do observe some evidence of ectopic expression from the BAC , particularly in the ventral nerve cord ( Figure S2; Figure S3 ) . Even with the slight level of ectopic expression , the Abd-B-Gal4 BAC seems to be a useful tool , as it recapitulates known patterns of Abd-B expression even in adult and larval tissues ( Figure S4 ) . Using this new reporter , we identified the adult male accessory gland ( see Figure 2B ) as a location of Abd-B expression ( Figure 2C ) . More specifically , based on the expression of our Abd-B-Gal4 BAC , Abd-B appears to be specifically expressed in the secondary cells ( Figure 2C , 2D , and 2E ) . To confirm this finding , we stained accessory glands in the presence of the Abd-B-Gal4 UAS-GFP reporter with an antibody directed against Abd-B ( Figure 2F , 2G ) . Like the reporter , accessory gland immunostaining against the Abd-B protein shows specific staining in the secondary cells ( Figure 2F , 2G ) . Interestingly , we also see Abd-B staining in the ejaculatory duct ( Figure S4C ) that is not observed with our reporter ( Figure 2C ) . This is perhaps not surprising , as the ejaculatory duct is a structure derived from the male genital disc , a tissue that primarily expresses the Abd-Br isoform [44] ( which is also recognized by our antibody ) . In order to examine the role of Abd-B in the development of the accessory glands , we sought a method to remove Abd-B expression exclusively in the secondary cells . Rather than use the traditional FLP-FRT system for making clones , we reasoned that in our collection of Abd-B cis-regulatory mutations [6] , we may already have a deletion that specifically removes secondary cell enhancers . Given our hypothesis that Abd-B might act as a cell fate determinant in the secondary cells , we screened a set of large , overlapping deficiencies covering the Abd-B cis-regulatory region for defects in secondary cell formation ( Figure 3A ) . To make this analysis easier , homozygous mutant flies were screened in lines that also contain a copy of our BAC reporter to mark the cells that would normally become secondary cells . Two of the lines examined , iab-6 , 7IH & iab-5 , 6J82 ( Figure 3C and 3D ) , showed a distinct morphological abnormality in the secondary cells . This abnormality can be easily seen using the cytoplasmic GFP marker . In wild-type cells , the GFP marker outlines the presence of large vacuolar structures in the secondary cells ( Figure 2D , Figure 3B ) . In both the iab-6 , 7IH & iab-5 , 6J82 mutants , these structures appear to be absent , and consequently the GFP marker is almost uniformly distributed across the cytoplasm . Although these secondary cells are not normal , we do not detect any expression of main cell-specific markers in these cells , suggesting that they are not transformed towards a main cell fate ( they still express the Acp95EF lacZ reporter gene [28] and fail to express the SP lacZ reporter gene ( data not shown ) [29] ) . To test if Abd-B is capable of transforming main cells into secondary cells , we expressed Abd-B across the whole accessory gland using a paired-Gal4 driver [46] . The most common result of this ectopic expression is cell death in the main cells ( data not shown ) . These results suggest that Abd-B expression in the secondary cells is required for morphological differentiation but may not be necessary for the initial differentiation between the two cell types . Based on the sequences uncovered by both the iab-6 , 7IH & iab-5 , 6J82 mutations , we concluded that the iab-6 domain , responsible for Abd-B expression in segment 6 , is also responsible for Abd-B expression in the secondary cells . Thus , we screened our collection of smaller iab-6 deficiencies [41] for the secondary cell phenotype . From this analysis , we were able to narrow down the location containing the secondary cell enhancer to a 2 . 8 kb region in iab-6 ( Figure 3A , 3F and 3H ) . Flies lacking this 2 . 8 kb region ( iab-6Δ5 ) specifically lack Abd-B protein expression in the secondary cells ( Figure 3H ) , and show distinct secondary cell morphological defects ( Figure 3F ) . Like the larger deficiencies above , iab-6Δ5 homozygous males lack the large vacuoles characteristic of secondary cells . As further confirmation of the importance of Abd-B and the 2 . 8 kb iab-6 enhancer in secondary cell development , we performed a number of control experiments . First , we crossed in a BAC transgene containing the wild-type Abd-B region and tested for rescue of the cellular phenotype . As expected , the secondary cells of males , homozygous for the iab-6Δ5 mutation but carrying one copy of the Abd-B BAC are substantially rescued ( containing a number of large vacuoles ) ( Figure S5C ) . Although this rescue is quite evident , it is not complete , a fact that probably reflects a weaker level of expression from the BAC relative to the native Abd-B locus . Indeed , Abd-B staining experiments using this BAC indicate that this is the case ( data not shown ) . Next , we created a transgene carrying the 2 . 8 kb region of iab-6 ( called D5-Gal4 ) and showed that it drives expression of Gal4 in the male reproductive tract specifically in the secondary cells ( Figure 3E ) . Using this D5-Gal4 driver , we were then able to drive expression of an Abd-B RNAi construct in the secondary cells . Knocking down Abd-B in the secondary cells was able to partially phenocopy the iab-6Δ5 mutation ( Figure S6C and S6F ) . The strength of this phenotype could be enhanced by the inclusion of a Dicer 2 overexpression transgene in the background . iab-6Δ5 was originally isolated in Iampietro et al [41] , where they did not observe any visible external phenotype . With the discovery of the secondary cell phenotype and the strong reproductive phenotype described below , we have renamed this allele iab-6cocu ( “cocu” means “cuckold” in French , reflecting that the mates of these males fail to reject other suitors ) . Interestingly , although the secondary cell enhancer was found in the iab-6 domain , it does not seem to be regulated like other BX-C enhancers . Previous work has demonstrated that most enhancers in the BX-C function coordinately through their integration into segment-specifically activated chromatin domains [6] , [47] , [48] . A special domain control element , called an initiator , is thought to dictate the activity state of a domain along the A-P axis [41] , [49] , [50] . Thus , deletion of the iab-6 initiator is predicted to inactivate Abd-B expression in the secondary cells , because the secondary cell enhancer should be coordinately regulated with the other enhancers in the iab-6 domain . In contrast to this prediction , we observed that deletions of the iab-6 initiator , which seem to show complete transformations of A6 to A5 , display wild-type accessory glands ( Figure 3E and 3G ) . From these experiments , we conclude that Abd-B expression in the secondary cells is set up by a different mechanism than that of tissues arising early in development . The iab-6cocu mutation offers the opportunity to investigate the role of the secondary cells in the PMR . First , we tested whether the main cells of these males are functional , since loss of main cell derived Acps may mask any secondary cell related phenotypes present in our mutant . We performed Western blots to examine the presence of known main cell Acps in the accessory glands of iab-6cocu males relative to two types of control males [males heterozygous for the iab-6cocu mutation ( henceforth referred to as control males ) and wild type males ( Canton S ) ] . As a negative control , we included the accessory glands of DTA-E males , which lack protein production in the main cells [19] but have theoretically normal secondary cells . We used antibodies to four Acps expressed in the main cells: SP [29] , Acp62F , Acp36DE [51] , and ovulin; the latter Acp is also present in the secondary cells but is known to be absent in DTA-E males [52] . We detected all four Acps in the extracts from iab-6cocu males ( Figure 4A ) . This result suggests that the main cells in iab-6cocu males are functional . To test if the iab-6cocu mutation impacts the ability of males to induce egg-laying in their mates , we crossed iab-6cocu males , control males , and DTA-E males to virgin females . During the first 24 hours after the start of mating ( ASM ) , the number of eggs laid by females that had mated to iab-6cocu males is comparable to that of mates of control males and is significantly higher than the number laid by females mated to DTA-E males ( Figure 5A ) . This indicates that the iab-6cocu mutation does not impact the STR and supports the Western blot results that suggest that the main cells are normal . However , egg laying in mates of iab-6cocu males decreased dramatically at 48 hours , and the total number of eggs produced over the entire 10 day period was significantly lower than the number laid by mates of control males ( Figure 5A ) ( note that Canton-S males behave similarly to our control males ( Figure S7 ) ) . This drop in egg laying is consistent with that observed when females do not receive or fail to store/release SP . This suggests that products from the secondary cells may be necessary for maintenance of the LTR . The proportion of progeny that eclosed from eggs laid by females mated to either iab-6cocu or control males was comparable , suggesting that there is no effect of secondary cell products on hatchability ( Figure 5B ) . Together these results suggest that the secondary cells perform a function that is essential for the maintenance of the post-mating egg-laying increase , but does not impact hatchability , and that this function is perturbed in iab-6cocu males . Under normal conditions , mated females are less receptive to subsequent mating for more than four days after the initial mating occurred [53] . This reduction in receptivity requires the receipt of Acps [19] , [23] , [26] . To test whether the iab-6cocu mutation alters female receptivity to remating , we mated virgin females to either iab-6cocu or control males and then allowed these females access to a single WT male at 1 d , 4 d , or 10 d ASM . At 24 hours after the initial mating , neither group of females remated , further suggesting that the STR is intact in mates of iab-6cocu males . However , when mated females were introduced to a WT male at 4 days ASM , females which had initially mated to iab-6cocu males were significantly more receptive than mates of control males . At 10 days ASM , both groups were fully receptive ( Figure 5C ) . Our results show that sexual receptivity is initially repressed in mates of iab-6cocu males , but that this effect is not maintained . This finding demonstrates a defect similar to those observed in known LTR-related proteins and further corroborates the LTR phenotype observed in our egg-laying experiments . To verify that both the egg-laying and receptivity phenotypes are caused specifically by the loss of Abd-B expression in the secondary cells , we also performed these experiments using BAC rescued iab-6cocu flies ( Figure S5 ) and D5-Gal4::Abd-B RNAi flies ( Figure S6 ) . In both cases , these experiments confirm a role for Abd-B in producing the PMR phenotypes . Again , as with the cellular phenotypes mentioned above , both the rescue and phenocopying was incomplete , though clearly significant . Given the caveats involved in these experiments regarding the level and timing of protein expression , this was perhaps not unexpected and demonstrate a strong relationship between the celluar and the behavioral phenotypes . Nonetheless , these data clearly point to a major role for Abd-B expression in the secondary cells in modulating the PMR . Our results suggest that the secondary cells are necessary for the processes required for long-term PMR maintenance . Therefore , the iab-6cocu mutation likely impacts proteins required for the LTR . While the Acps that are produced and transferred to females have been extensively described , [51] , [54]–[57] little is known about their cellular origins . Thus , we investigated the possibility that some of the known PMR-related Acps , and more specifically those involved in the LTR , could be produced in the secondary cells or in both cell types , and thus , may be absent in iab-6cocu males . We performed Western blots to examine the presence of known LTR Acps in the accessory glands of iab-6cocu males relative to control males . As a negative control for main cell expressed Acps , we included DTA-E males . Any secondary cell-expressed Acp should still be produced in these males , but main cell expressed Acps should not . We used antibodies to six Acps that regulate the PMR; one STR associated Acp ( CG11864 ) and five LTR associated Acps ( CG10586 ( Seminase ) , CG9997 , CG17575 , CG1656 , and CG1652 [36] , [38]–[40] ) . All of these Acps were present in iab-6cocu males ( Figure 4B ) . Surprisingly , three of the Acps associated with the SP pathway ( CG17575 , CG1656 , and CG1652 ) were also present in AG extracts from DTA-E males suggesting that they may be secondary cell expressed . Supporting this hypothesis , RNAi of these proteins in the secondary cells knocksdown their expression , while leaving a main cell-expressed control protein , Acp62F , unchanged ( Figure S8 ) . The remainder of the Acps , CG9997 , CG11864 , and Seminase , are likely all expressed primarily or exclusively in the main cells . Further , these results support our previous conclusion that the secondary cells in iab-6cocu males maintain a distinct expression profile from main cells and still produce some secondary cell proteins . The LTR defects seen in mates of iab-6cocu males are consistent with those associated with failure to store or release SP [51] , [55]–[57] . However , iab-6cocu males produce SP and all known LTR Acps . Still , it is possible that the iab-6cocu mutation interferes with the ability of SP to enter storage and thus maintain the LTR . We tested for this by performing Western blots using SP antibodies . Both control and iab-6cocu males transfer SP to their mates , and there are comparable amounts of SP in the female reproductive tract by 2 h ASM . However , by 24 hours ASM and continuing to 6 days thereafter , mates of iab-6cocu males have significantly less stored SP ( Figure 4C; Figure S9 ) . To distinguish between premature loss of SP from the seminal receptacle ( SR ) versus failure of SP to be stored in the SR initially , we performed Western blots of SRs dissected from females mated to either iab-6cocu or control males . Significantly less SP is present in the SR of mates of iab-6cocu males at 2 h ASM compared to mates of control males , while the amount of SP present in the remainder of the reproductive tract is comparable , though slightly higher in mates of iab-6cocu males ( Figure 4D; Figure S9 ) . These results suggest that iab-6cocu males transfer normal amounts of SP but that this SP fails to enter the SR . The reduction/absence of stored SP at later time points ( 1–7 days ASM ) is likely responsible for the reduction in egg laying and the increase in receptivity seen in mates of iab-6cocu males . SP also plays a role in sperm competition [58] , which occurs when ejaculates from two males are present within the same female reproductive tract [59] . For example , in circumstances where SP null males are the first male to mate with a given female , they sire a higher percentage of the total progeny ( P1 , #progeny from first male/total progeny ) than control males [33] . To test whether the iab-6cocu mutation also affects P1 , we mated iab-6cocu and control males to cn bw females and , after 3 days , allowed them to mate with cn bw males . The iab-6cocu males had significantly higher P1 than control males consistent with a problem in SP presence or storage ( Figure 6A ) . One possible explanation for the reduction in stored SP in mates of iab-6cocu males is a defect in sperm entry into storage or an increase in the rate at which sperm are released from storage . To test this , we counted sperm present in both female sperm storage organs at 2 hours , 4 days , and 10 days ASM . Females mated to either control or iab-6cocu males store sperm at comparable levels at 2 hours ASM and appear to retain normal numbers through 4 days ASM ( Figure 6B and 6C ) . These results suggest that initial sperm storage and release is normal in mates of iab-6cocu males and that the reduced level of SP in the seminal receptacle at 2 hours ASM is not due to a failure to adequately store sperm . Mates of iab-6cocu males do not show the stereotypical sperm over-retention phenotype seen with knockdown of other LTR related proteins [33] and instead show a slight but significant decrease in stored sperm within the SR at 10 days ASM when compared to controls ( Figure 6B ) . This is not wholly surprising , as the iab-6cocu mutation does not result in the absence of a single gene product , but likely several that contribute to different aspects of the PMR . It is possible that the secondary cells produce , modify , or transfer some product necessary for sperm to be retained within the female sperm storage organs . When SP is absent , this product may be regulated improperly resulting in the typical sperm over-retention phenotype . A loss of , or reduction in this product , combined with the SP retention defect , may explain these results . Our Western blots and RNAi data showed that three of the known LTR specific proteins ( CG1656 , CG1652 , and CG17575 ) are produced in the secondary cells and that one ( CG9997 ) is likely produced by main cells ( Figure 4B and Figure S7 ) . We considered the possibility that failure to transfer one or all of these Acps to the female during mating could contribute to the SP storage defect seen in mates of iab-6cocu males . To test this , we performed Western blots on the reproductive tracts of females mated to either iab-6cocu or control males at 15 m , 30 m , and 1 h ASM using antibodies to CG9997 , CG1656 , CG1652 , and CG17575 . Although all four Acps are transferred to females and are present in the female reproductive tract at all time points tested , their abundance , gel mobility , or processing appear to be abnormal in mates of iab-6cocu males ( Figure 7 ) . Both CG1656 and CG1652 run at a lower apparent molecular weight in iab-6cocu males compared to control males ( Figure 7A , Figure 4B ) . Ovulin , likewise , shows reduced apparent molecular weight in iab-6cocu males ( Figure 7C , Figure 4A ) . The gel mobility differences for these proteins in iab-6cocu versus control males is evident both within AG extracts as well as across time points ( 15 m , 30 m , and 1 h ASM ) within the female reproductive tract of their mates . Ovulin is normally processed inside the female reproductive tract [60] . This processing appears to occur properly in mates of iab-6cocu males , although the apparent molecular weight of some cleavage products is altered . It is unlikely that these differences in apparent molecular weight are caused by sequence differences or background effects . The controls used for these experiments are heterozygous for all of the chromosomes in the iab-6cocu mutant line . Thus , if the gel mobility differences were caused by sequence differences , we would expect to see two bands indicating the WT and altered version of each protein . Further , ovulin and CG1656/1652 are located on separate arms of chromosome 2 and are necessary for different aspects of the PMR . Together , these observations suggest that the gel mobility differences may be a result of posttranslational modification . Ovulin is a glycoprotein [52] , but little is known about the posttranslational modifications of CG1656 and CG1652 . To test whether differences in glycosylation underlie the gel mobility differences observed , we treated extracts from control and iab-6cocu males with PNGaseF . This treatment , which removes N-linked glycosylation [61] , [62] , resulted in loss of the apparent molecular weight differences between iab-6cocu and control flies for all three proteins ( Figure 8 , CG1652 not shown ) . These results suggest that the secondary cells contribute to the regulation of posttranslational modifications , and more specifically glycosylation , of Acps . They are also the first evidence of the presence of N-linked glycosylation on CG1656 and CG1652 . To verify that the glysocylation differences are caused specifically by the loss of Abd-B expression in the secondary cells , we also looked at the gel mobility of CG1656 and CG1652 in BAC rescued iab-6cocu flies ( Figure S5 , CG1652 not shown ) . The rescue BAC was able to restore proper glycosylation in the AGs of some males but not others . This is consistent with the partial rescues we have observed , especially if the glycosylation phenotype is dose dependant and the BAC males are on the threshold . Still , these results support the connection between Abd-B expression in the secondary cells and proper glycosylation of Acps and suggests that there may be a connection between these glycosylation differences and the PMR . In wild type males , CG9997 is transferred to females as a full length protein ( 45 kD ) and is processed in the female reproductive tract to a smaller form ( 36 kD ) . Both products persist in the female for longer than 1 h . Males with the iab-6cocu mutation produce full length CG9997 but the full length form does not persist inside the female reproductive tract ( Figure 7A ) . A similar increase in processing or instability of the full length product was seen in males that do not produce or transfer CG1656/CG1652 [39]; its biological relevance is currently unknown . Since iab-6cocu males produce and transfer CG1656/CG1652 , our results suggest that in addition to these two proteins , the products of the secondary cells are essential for regulating the stability of CG9997 inside the female . As observed with the other LTR Acps we assayed , CG17575 is produced and transferred to females by iab-6cocu males . However , CG17575 is present in higher amounts in the reproductive tract of females mated to iab-6cocu males at all time points when compared to controls ( Figure 7B; Figure S9 ) . Whether this difference is due to increased transfer or failure to degrade CG17575 within the tract is unclear .
Due to the large size and complexity of the Abd-B regulatory region , we created a BAC- reporter construct to monitor Abd-B expression in the adult fly . When combined with fluorescent markers , this method allowed us to bypass the technical issues of antibody penetration and the laborious dissections needed for in situ hybridization or immunohistochemistry to identify a novel area of Abd-B expression in the adult . Overall , the BAC reporter is able to accurately reproduce the known , complex Abd-B expression pattern; indeed , our BAC construct seems to more-faithfully reproduce Abd-Bm expression than even a previously isolated transposon insert in the Abd-B promoter ( Abd-B-Gal4LDN ) [63] . Furthermore , by combining our BAC reporter with pre-existing deletion mutations , we were able to discover the function of a vital gene in an adult tissue without the need to create mitotic clones . From the standpoint of Hox gene regulation , our discovery of the secondary cell enhancer is quite interesting because , unlike other cell-type specific enhancers from the BX-C , the secondary cell enhancer does not seem to be regulated by a domain initiator [6] , [41] , [49] . Most cell-type specific enhancers from the BX-C are not intrinsically restricted along the A-P axis . They are restricted only to a specific cell-type and gain A-P restriction through clustering in a BX-C domain . For example , in a transgene assay , an Abd-B enhancer from the iab-7 domain ( called 11X ) drives expression in the tracheal placodes in all segments . However , in the BX-C , this enhancer seems to be active only in the Abd-B expression domain [6] . The clustering of enhancers to one area of the chromosome is thought to allow all of the enhancers to be coordinately regulated along the A-P axis as a domain through the changing of the local chromatin environment . The Polycomb ( Pc ) repression machinery is thought to be critical for this process by creating repressive chromatin over inactive domains ( [6] and refs . therein ) . Specialized elements , called initiators , seem to read an A-P segmental address and act as domain activators , probably by preventing Pc repressive complexes from establishing on active domains [41] . The domain model predicts that deletion of an initiator element should prevent domain activation , leaving all enhancers in its domain inactive [41] . Based on this paradigm and the fact that the secondary cell enhancer was found in the iab-6 domain , we expected that the deletion of the iab-6 initiator would abolish Abd-B expression in the secondary cells . However , we found that Abd-B expression in the secondary cells of initiator mutants was normal . This finding argues against the strictest interpretation of the initiator model . We can propose several hypotheses to resolve this discrepancy . For example , the Pc repression system is known to act on many genes during development , but its main targets appear to be the homeotic genes during the establishment of segment identity . It is possible that late in development , after cells have made initial cell fate decisions ( and the segment identity role of the homeotic genes might be less important ) , the targets of Pc silencing might change . Such loosening of Pc silencing over the Abd-B cis-regulatory regions would allow previously silenced enhancers to become available for regulating Abd-B expression so that it could perform other functions , such as in the secondary cells . Alternatively , the difference in Abd-B gene regulation that we observe in secondary cells may reflect the cellular origin of the secondary cells . Most BX-C cis-regulatory mutations were isolated based on cuticular phenotypes and confirmed using antibody staining in the epidermis and CNS . These tissues are of ectodermal origin , unlike the mesodermally-derived secondary cells [64] . Perhaps , the rules governing the coordination of Hox expression in the ectoderm do not hold true in the other germ layers . Consistent with this , BX-C genes are expressed differently in the gut visceral mesoderm than in the ectoderm [65] . Evolutionary considerations may provide some explanation for why the fly uses different means of controlling Abd-B expression in embryonic segment identity specification vs . in later reproductive tissues . Abd-B class Hox proteins play roles in the formation of the external genitalia in both arthropods and mammals [66]–[70] . Due to the similarity in expression pattern and function , it has been proposed that Abd-B's role in the formation of the genitalia predates its role in segmental identity [67] [71] . Here , we have shown that Abd-B is also important for correct development of cells within the Drosophila male AG that produces many seminal fluid proteins required for male reproductive success . The mammalian orthologues of Abd-B , the Hox9 to 13 class of genes , are expressed in the developing seminal vesicle and prostate gland , both seminal protein secreting organs [72] , [73] . The analogy in function between these organs , and their similarity in gene expression patterns suggests that the role of Abd-B class genes in the male reproductive tract might be an ancient , conserved function , potentially independent of its role in segmental identity . In this light , it would not be surprising that Abd-B regulation in the secondary cells escapes the domain regulation seen for Abd-B function in segment identity determination . The cis-regulatory domains for segment identity could have been added separately , possibly through co-opting the abd-A gene regulatory regions , as suggested by transvection studies [74] , [75] . In any case , the adding of cis-regulatory sequences and the consequent constraints of the domain model on Abd-B would necessarily have to preserve its late function in the secondary cells . Previous studies have shown that the male AG produces Acps required to initiate and maintain a range of PMRs in females . Further , diptheria-toxin mediated-ablation of accessory gland main cells demonstrated that products of these cells are essential for the PMR . The importance of the main cells was further strengthened by the discovery that they produce the Acp sex peptide ( SP ) , which is essential for many aspects of the PMR and whose persistence allows maintenance of PMR effects ( i . e . the LTR ) [29] . Additional Acps important for other aspects of the PMR ( e . g . Acp36DE , ovulin ) were also found to be produced by main cells . Thus , until now , the role of the secondary cells was unknown and no methods to directly target these cells were available . Using the iab-6cocu regulatory mutant of Abd-B , we demonstrated that the secondary cells make a unique contribution to maintenance of the female's post-mating changes in egg-laying , receptivity , sperm competition , and sperm storage ( summarized in Figure 9 ) . Our results are consistent with findings about secondary cell function obtained independently by Minami et al . , from their study of dve mutants [76] . The inability of iab-6cocu males to maintain the PMR in their mates is consistent with perturbation of the function of the “LTR network” . These LTR network proteins are needed to promote the association of SP with sperm in the SR , an association that is required for SP-mediated maintenance of the PMR [22] , [39] . The results of our study also show that three of these LTR network Acps , CG1656 , CG1652 , and CG17575 are all produced in the secondary cells while CG9997 and Seminase are primarily or exclusively main cell expressed . This is the first direct evidence that Acps from both cell types work together in a complex pathway . While all reported LTR specific network proteins ( CG9997 , CG1656 , CG1652 , and CG17575 ) are present in the iab-6cocu mutant , they are all abnormal , either in amount/stability inside the female reproductive tract or in glycosylation state; how these changes result in the SP storage defect is an important area for future study . The phenotype of iab-6cocu males shows one exception to the standard LTR phenotype . Mates of SP null males ( or males knocked down for any of the other 4 LTR related proteins ) have high rates of sperm retention; this phenotype is lacking in mates of iab-6cocu mutants . The mechanisms behind the release of sperm from storage and this sperm retention phenotype are unknown . Loss of Abd-B expression most likely impacts a wide variety of Acps and potentially cellular functions associated with vacuoles ( such as transmembrane transport ) . Thus , it is possible that one or a combination of the proteins affected by this mutation ( some as yet unidentified ) may negatively impact normal sperm storage independent of the influence of Abd-B on the storage of SP in the seminal receptacle . This could result in masking the over-retention of sperm seen in SP nulls due to the loss or abnormal function of a protein or proteins necessary for the retention phenotype to occur . Further work investigating the role of individual secondary cell associated Acps in the LTR , as well as sperm storage , may be helpful in determining how these processes function . The iab-6cocu mutant revealed another unique role for secondary cells: regulating glycosylation of at least three seminal proteins ( ovulin , CG1656 , and CG1652 ) ; the impact of glycosylation on the functions of these proteins is as yet unknown . Our findings show that CG1656 and CG1652 are produced by the secondary cells . However , ovulin is produced in main cells as well as secondary cells , and is detected in the AG lumen as well as in the vacuoles of the secondary cells [52] . As such , we suggest three possibilities for how secondary cells might mediate glycosylation: 1 ) through the secretion of glycosylation substrates that can be taken up and used by main cells , 2 ) through the secretion of glycosyation regulators directly into the lumen where they could modify Acps from both cell types that are present there , or 3 ) that Acps like ovulin and other main cell products are taken up into these vacuoles and then modified before being secreted into the lumen as mature , glycosylated proteins . Future dissection of the glycosylation phenotypes in the iab-6cocu mutant males will help shed light on the role glycosylation plays in regulating the PMR and how this process is regulated in the tissue as a whole . It is important to note that our Abd-B regulatory mutant still has secondary cells , or at least precursor cells poised to become secondary cells . The initial differentiation step that allows for Abd-B expression in the secondary cells is not perturbed by the iab-6cocu deletion . As such , although we have already found a new and unanticipated role for secondary cells in regulating the PMR , it is possible that these cells play additional roles . Future cell ablation experiments using tools derived from this study will allow tests of such additional roles . In conclusion , we have shown that each of the two cell types of the Drosophila AG plays important roles in producing the female PMR: the main cells producing several Acps to initiate and maintain the PMR , and the secondary cells providing products to aid in temporally extending the response . We expect that through the action of various combinations of transcription factors , like Abd-B , the two cell-type lineages have diverged into distinct , and specialized cell-types . Although these two cell types perform vitally intertwined functions , they are maintained as separate cell types . This suggests that there may be a requirement for compartmentalization of their functions or products , or that the two cell types evolved separately for some other purpose and were functionally associated afterward . It is , therefore , of great future interest to identify the specific products of each of these cell types and to determine how they work in conjunction to mediate the full reproductive effect of seminal secretions .
The Abd-B rescue BAC was constructed from BACR24L18 ( size - 171936 bp; AC095018 ) , which contains the Abd-B region of the BX-C . We first constructed a pW25-based vector [77] to be used to add sequences to the BAC for site-specific integration into the Drosophila genome . The original white gene carried by pW25 was replaced by a Su ( Hw ) insulator-flanked white gene from the SUPor-P plasmid [78] by amplifying it with NotI-forward and AscI-PmeI reverse ( see Table 1 for primer list ) . These primers carry restriction sites for insertion of the product into pw25 after appropriate enzyme digestion . A fragment containing the kanamycin resistance gene ( KanR ) was amplified from pIGCG21 [79] using the following primers: PmeI5′Kan , 5′attB3′KanAS . Next , an AttB sequence was PCR amplified using the primers 3′Kan5′attBS , and PmeI3′attBAS . The KanR and AttB fragments were then mixed together with the PmeI5′Kan and PmeI3′attBAS primers for a final overlap PCR reaction . The resulting PCR fragment , containing the KanR gene and an AttB sequence flanked by PmeI sites , was cloned into pGemTeasy . After sequencing , this fragment was excised with PmeI and cloned into the unique PmeI site of the modified pW25 vector ( resulting in pW25/Kan-AttB ) . To mediate recombination in bacteria , two homology regions were then added to this vector . First , an 859 bp fragment from the iab4 region ( iab4 HR: 101409–102267 coordinates in the BACR24L18 ) was made by PCR with the primers: NotI iab4 and EagI iab4 . The resulting PCR fragment was cut with NotI and inserted into the unique NotI site of the modified pW25 . A second homology fragment was designed to target the SacBII gene present on the BAC backbone . Using the PCR primers: MluI SacBII and AscI SacBII , a 525 bp fragment from SacBII was amplified ( SacBII HR: 921–1445 coordinates in the BACR24L18 ) . The resulting 525 bp fragment was digested with AscI and MluI , and inserted into the unique AscI site of pW25/Kan-AttB . Clones for both homology regions were selected in an orientation required by the homologous recombination process to function correctly . The completed construct was digested with the ISce-I endonuclease and the fragment containing the homology regions flanking the KanR gene , the white gene and the AttB site was gel purified . This fragment was then used to recombineer BAC24L18 using the protocol of Soren Warming [80] . The resulting BAC , called iab4-SacBII BACR24L18D ( 108528 bp ) , contains the region of the BX-C from about iab-4 to the Abd-B m promoter . The overall structure of this BAC was verified by restriction enzyme mapping [using three restriction digests ( EcoRI , XmaI , BamHI ) ( data not shown ) ] . Using iab4-SacBII BACR24L18D as a base we used recombineering to replace the start codon of the Abd-Bm isoform with sequence encoding the Gal4 transcription factor . First , a negative/positive selection cassette was created using the SacB gene and the Ampicillin resistance ( AmpR ) gene . SacBII was digested out of the plasmid pSK2-SACBK MAR using BamHI and EcoRI and cloned in pHSS7 [81] . The AmpR gene was then amplified with primers ( Amp BamHI and Amp XmaI ) carrying a BamHI and XmaI site . The amplified AmpR fragment was digested with BamHI and XmaI and cloned into a pHSS7-SacBII digested with the same enzymes , producing pHSS7-SacBII/Amp . The SacBII/Amp cassette was flanked by two large homology regions using an overlap PCR strategy , as follows: The primers NHR L frw NotI/long and NHR L rev OL/long were first used to amplify the region 39201 bp to 40325 bp of iab4-SacBII BACR24L18D . At the same time , the primers NHR R frw OL/long and NHR R rev XmaI/long were used to amplify the 40326 to 41590 bp region of iab4-SacBII BACR24L18D . The two reaction products contain a 27 bp region of complementarity to mediate overlap PCR . Thus , the two fragments were mixed together with the primers NHR L frw NotI/long and NHR R rev XmaI/long in an overlap PCR reaction . The resulting fragment , containing the two homology domains fused together , was digested with NotI and XmaI and cloned into pHSS7 ( pHSS7-NHR-L/NHR-R ) . Next , from the previously created pHSS7-SacBII/Amp , the double selection cassette was amplified with primers NSx/A new F EcoRI and NCSx/A new R XbaI . Digesting this PCR fragment with EcoRI and XbaI produced a fragment that could be inserted between the two homology domains of pHSS7-NHR-L/NHR-R creating pHSS7-NHR-L/SacBII-Amp/NHR-R . The primers F300 NS/A rec and R300 NS/A rec were used to amplify a fragment from pHSS7-NHR-L/SacBII-Amp/NHR-R for recombineering . This fragment contained 300 bp of the homology regions on both sides of the cassette carrying the SacBII and AmpR genes . Once again recombineering was performed using the protocol of Soren Warming [80] , where the target BAC was iab4-SacBII BACR24L18D . BAC DNA purified from the resulting ampicillin resistant/sucrose sensitive colonies were verified by extensive restriction enzyme digests . The new BAC was named iab4-SacBII BACR24L18D N S/A ins . To replace the SacBII/AmpR cassette , the Gal4 gene ( with a synthetic polyA tail ) was PCR amplified from the plasmid pTnT Gal4 ( unpublished , pTNT base vector from Promega Corp . , Madison , Wisconsin , USA ) using the following primers: HR-R Gal4rep and HR-L Gal4rep . These primers contain 55 bp homology regions to mediate recombineering to the BX-C sequences just flanking the SacBII/AmpR cassette . The resulting targeting fragment was then phosphorylated by T4 kinase in order to improve the recombination reactions . After standard preparation of the recombineering DY380 strain containing iab4-SacBII BACR24L18D N S/A ins , bacterial colonies were selected on LB agar plates containing 10% sucrose . Restriction digestion of the candidate colonies with BamHI was performed in order to confirm the correct replacement of the SacBII/Amp cassette with Gal4 . Using the PhiC31 system ( [45]; www . flyc31 . org ) , site 51C on the second chromosome was chosen for integration of the BACs into the fly genome . For better integration frequencies , all BACs were isolated on the day of injection using the NucleoBond PC 20 ( Macherey-Nagel ref 740571 ) miniprep kit and resuspended in injection buffer [82] . Embryos were injected with BAC DNA ( at about 50–100 ng/ul ) through the chorion using the Eppendorf system ( FemtoJet & TransferMan NK 2 ) equipped with Femtotips II glass needles . Integration efficiency was about 5% , based on the total number of fertile adults that yielded at least one integrant . The 2 . 8 kb putative enhancer sequence removed in iab-6cocu was amplified by PCR using primers D5 F and D5 R . Both of these primers contain a BamHI site at their 5′ ends . The amplified DNA fragment ( called D5 ) was cloned into the BamHI site of the pChs-Gal4 plasmid , which contains a minimal Hsp70 promoter upstream of the coding sequence for Gal4 and the HSP70 3′UTR ( Drosophila Genomics Resource Center [83] . Although clones with the enhancer in both orientations were isolated , we proceeded using a clone where the D5 R primer containing end was closest to Gal4 . The D5-Gal4 cassette was then digested out of the pChs-Gal4 vector with NotI and cloned into the NotI site of pattB [45] . An insertion with the Gal4 coding sequence next to the white gene was selected for injection . This construct was integrated by Genetic Services Inc ( Cambridge , Mass ) into the VK00001 ( 59D3 ) platform [84] . The resulting integrant is named D5rsG4rs and referred as to D5-Gal4 in the text . All crosses were done using standard genetic techniques . iab-7Sz , iab-6 , 7IH , iab-5 , 6J82 , and iab-4 , 5 , 6DB are described in [6] . The lines iab-6Δ5 and iab-64 are described in [41] . The line iab-6Δ5 was described as a deficiency without any phenotypic consequence . Following the LTR phenotype identified in this work the line was renamed iab-6cocu ( reflecting that mates of these males fail to reject other suitors; “cocu” means “cuckold” in French ) . The BAC-AbdBGal4 , w+ , UAS-GFP/Cy line carrying the Abd-B Gal4 BAC reporter and a UAS-GFP marker on the second chromosome was created for this study by recombining a chromosome carrying the BAC and a UAS-GFP chromosome . The BAC reporter chromosome cannot exist as a homozygote . The 4 . 4E transgenic lacZ reporter line is described in [6] . The Gal4 expressing lines driven by a paired enhancer ( ( w−; prd-mf5 . 2 , w+/CyO ) , ( w−;prd-mf5 . 4 , w+ ) , ( w−;prd-mf5 . 5 , w+ ) , ( w−;prd-mf9 . 3 , w+ ) , ( w−;prd-mf9 . 7 , w+ ) ) were obtained from Makus Noll's laboratory [46] . They were used in a cross with a UAS-AbdBm [85] flies for the experiment in which we tested for the ability of Abd-Bm to transform main cells into secondary cells . All flies for fertility and fecundity assays , tests of receptivity and sperm competition , Western blotting , sperm counts , and PNGase F assays were raised at room temperature ( 23±1°C ) . Females were aged 3–5 days from eclosion in groups of 7–11 in glass vials on standard yeast-glucose media with added yeast . Males were aged 3–5 days from eclosion in groups of 10–20 in glass vials on standard yeast-glucose media . Antibody and X-Gal staining on embryos and dissected accessory glands was performed as described in [12] and [42] respectively , using a 20 min fixation . The Abd-B primary antibody , obtained from the Developmental Studies Hybridoma Bank , was diluted 1∶4 . Goat-anti-mouse secondary antibody , coupled to Alexa Fluor 488/555 X ( Invitrogen AG ) , used to reveal Abd-B localization was used at 1∶500 dilution . Goat HRP coupled anti-mouse was obtained from Biorad and used at 1∶1'500 dilution . Staining with FM4–64 dye was done by placing a drop of the dye onto a microscope slide and placing a freshly dissected gland into it . The glands were immediately covered with a cover slip and visualized using fluorescent microscope at 555 nm . In all assays , we used 3–5 day old virgin females from a wild type strain ( Canton-S for fertility/fecundity assays , receptivity assays , sperm counts , and for Western blotting experiments; and cn bw for sperm competition assays ) . Females were placed singly in glass vials with food and allowed access to an iab-6cocu , control male ( heterozygous for the iab-6cocu mutation ) , DTA-E , or WT Canton-S male . Pairs were watched to confirm that mating had occurred . The male was removed upon dismounting . All statistical analysis was performed with the Jmp9 software . In the fertility/fecundity assays , after mating , individual females were housed for 24 hours in glass vials on yeast-glucose media . After 24 hours each female was transferred to a fresh vial , and the eggs laid in the previous vial were counted . This process was repeated for a total of 10 days . Upon eclosion , all progeny from each vial were counted . Hatchability ( # progeny/# eggs ) was calculated per day and across the 10-day period for each female . Values greater than 1 represent instances where the number of progeny produced exceeded the number of eggs observed . This is an accurate representation of counter error , and was not normalized to 1 . Small levels of counter error has a greater impact on hatchability for females that lay few eggs . Comparisons of egg and progeny production between control and experimental females were performed using a Wilcoxon non-parametric test and statistics comparing the overall 10 day trends were performed using a repeated measures ANOVA . | Similar to the prostate gland and seminal vesicle in mammals , the Drosophila male accessory gland produces seminal fluid proteins that are critical for individual male reproductive success . In Drosophila , many of these proteins function to induce a suite of long-lasting physiological changes in mated females , like increased egg laying and decreased receptivity to secondary courtships . While investigating the cis-regulatory region of the Hox gene , Abd-B , we found that Abd-B is specifically expressed in a mostly uncharacterized cell-type of the accessory gland , called the secondary cells . Using regulatory mutants of Abd-B , we were able to perturb the function of the secondary cells to show that the secondary cells play a critical role in extending the duration of the post-mating response in females . The induced physiological changes in the female result in a genetic advantage for the genes of the copulating male in populating the next generation . Interestingly , the function of the Abd-B class genes in seminal protein producing tissues seems to be an ancient and conserved function , as the orthologues of Abd-B in mammals , the hox9-13 class of genes , are expressed in the mammalian prostate and seminal vesicle . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"genetic",
"mutation",
"fertilization",
"gene",
"expression",
"genetics",
"molecular",
"genetics",
"biology",
"pattern",
"formation",
"genetics",
"and",
"genomics",
"cell",
"differentiation",
"evolutionary",
"developmental",
"biology",
"gene",
... | 2013 | A Novel Function for the Hox Gene Abd-B in the Male Accessory Gland Regulates the Long-Term Female Post-Mating Response in Drosophila |
The phosphorylation of a substrate at multiple sites is a common protein modification that can give rise to important structural and electrostatic changes . Scaffold proteins can enhance protein phosphorylation by facilitating an interaction between a protein kinase enzyme and its target substrate . In this work we consider a simple mathematical model of a scaffold protein and show that under specific conditions , the presence of the scaffold can substantially raise the likelihood that the resulting system will exhibit bistable behavior . This phenomenon is especially pronounced when the enzymatic reactions have sufficiently large KM , compared to the concentration of the target substrate . We also find for a closely related model that bistable systems tend to have a specific kinetic conformation . Using deficiency theory and other methods , we provide a number of necessary conditions for bistability , such as the presence of multiple phosphorylation sites and the dependence of the scaffold binding/unbinding rates on the number of phosphorylated sites .
The multisite ( de ) phosphorylation system is modeled using a standard sequential mechanism ( Figure 1A ) . To introduce the scaffold we allow for reversible binding between the scaffold protein and the substrate with i phosphorylated sites , to form the complex ( Figure 1C ) . We allow phosphorylation to take place only for the scaffold-bound substrate , due to the fact that scaffolds accelerate substrate phosphorylation either by tethering the kinase and the substrate in proximity to each other , or by allosterically activating the kinase or the substrate [23] , [24] . The degree of rate acceleration by scaffold proteins can be as much as 10 , 000 fold [23] . With regard to dephosphorylation , it has been proposed that some scaffold proteins may protect bound proteins from the action of phosphatases [25] , [26] , while other scaffold proteins actually recruit phosphatases in addition to kinases [27] . We assume by default that dephosphorylation takes place equally on and off the scaffold , but we will also consider cases where phosphatases act exclusively off the scaffold . To quantify the dynamics of multisite phosphorylation , we have explored two types of commonly used mechanisms: full mass action kinetics ( MA ) [12] , [13] , [14] , and simplified linear enzymatic rates ( LR ) . In the linear rate model LR , the rates of flux of through phosphorylation and dephosphorylation are given by and respectively , where and are the total kinase and phosphatase concentrations ( Figure 1D , F ) . In the full model MA , the free kinase concentration is distinguished from the total kinase concentration , and phosphorylation follows a standard Michaelis-Menten mechanism of complex formation using , , and as the on , off , and catalytic rates , respectively . Similarly for the dephosphorylation mechanism ( Figure 1B , E , G ) . The full model has many more variables , parameters , and nonlinear reaction terms than the simplified LR model for a given total number of sites , which in practice means that LR is more amenable to mathematical analysis [16] . In fact , it is known that in the absence of a scaffold the LR model always results in a unique steady state , while the full model can support multistability [12] , [13] , [14] . We termed the simplified model without scaffold as “LR-NS” ( Figure 1D ) , the simplified model with scaffold as “LR-S” ( Figure 1F ) , the full model without scaffold as “MA-NS” ( Figure 1E ) , and the full model with scaffold as “MA-S” ( Figure 1G ) . It is worth pointing out that a distributive mechanism is assumed for ( de ) phosphorylation on scaffold , that is , that the enzymes tend to unbind from the substrate after each modification . There is evidence that some scaffold proteins may behave in this way . For example , the Ste5 scaffold protein binds weakly to its associated kinases [28] , and it has even been hypothesized that one of those kinases ( Ste7 ) may be frequently ejected from the Ste5 as a result of feedback phosphorylation [29] . Similarly , human MEK1 protein , when bound to the KSR scaffold protein , is thought to be phosphorylated by an ( unbound ) trans-acting homodimer of the RAF kinase [30] . If a kinase were to remain bound to the scaffold through multiple , processive phosphorylation events , however , this would be expected to reduce the propensity of the scaffold to promote bistability .
Before investigating the parameter patterns of bistable multisite ( de ) phosphorylation systems with scaffold , we first explore network topologies that exclude bistability regardless of kinetic parameter values . To this end , we employ the deficiency theory developed by Feinberg and others [20] , [21] , [22] , and we restrict our attention to the simplified linear rate model with scaffold , LR-S . The deficiency theory of chemical reaction networks is able to predict under certain circumstances that a given system is incapable of having multiple steady states , regardless of the parameter values used ( assuming fixed total protein concentrations ) . In order to do this , it only makes use of qualitative graphic-theoretic properties of the network , such as the number of connected components in the reaction diagram , and the number of nodes in this diagram , called complexes . For instance , the reaction network , has complexes ( , , , and ) and connected components . The deficiency of the network is defined as , where is the rank of the stoichiometry matrix . The most widely used result in the theory is the Deficiency Zero Theorem , which states that if and every connected component is strongly connected ( such as in the simple example above ) , then multistability is impossible , regardless of parameter values . That theorem is the basis for several of the results in this analysis . Please refer to Section 1 in Text S1 for details on the proofs of all results . If the scaffold association and dissociation rates and are independent of the phosphorylation state of the system , i . e . , and are constant for all values of , then there cannot be multiple steady states ( Figure 2A ) . In other words , to achieve multistability , the scaffold binding mechanism must be related to , or affected by , the phosphorylation state of the substrate . This is consistent with the finding in [16] that scaffold sequestration rates need to vary with the phosphorylation state , in order to affect the ultrasensitive behavior of the system . Proof sketch: define the variables , . Since the phosphorylation state is irrelevant for the scaffold binding and unbinding reactions , the variables , are the solutions of system . If phosphorylation and dephosphorylation only take place for scaffold-bound substrates , then the system can only have one steady state ( for given total concentrations of the substrate , enzymes and scaffold ) ( Figure 2B ) . The same conclusion holds if phosphorylation and dephosphorylation only take place away from the scaffold ( Figure 2C ) . In order to allow for bistability , both the scaffold-bound and the scaffold-unbound proteins must have access to at least one type of enzyme – the kinase or the phosphatase . Proof sketch: the two statements follow directly from the Deficiency Zero Theorem – however notice that Figure 2B and 2C are only diagrams in that the complex is shortened as . In Figure 2B , can be easily included as necessary and , , . In Figure 2C , including in the scaffold binding reactions but not the phosphorylation reactions forces to rewrite the graph as shown in Section 1 . 2 of Text S1 , and , , . We point out that even though the model in Figure 2C is always monostable , this particular topology has shown to be highly ultrasensitive for some parameter values [16] , which underscores the difference between ultrasensitive behavior and bistability . Even in the presence of a scaffold , a substrate with a single phosphorylation site is incapable of producing bistable behavior for several possible network configurations ( Figure 2D ) . This result provides evidence that if the kinase and the substrate both remain bound to the scaffold long enough , on average , for the kinase to catalyze two or more phosphorylation events in a processive manner , then the propensity for scaffold-driven bistability will be reduced . The proof for all configurations given in Figure 2D is given in Section 1 . 4 of Text S1 , and it is based on exploring the signs of the entries in the stoichiometric matrix as well as all its square submatrices . We take a closer look into the parameter values of the linear rate scaffold system LR-S , in search for patterns that might make bistability more likely . We simulate this system using a large set of phosphorylation , dephosphorylation , and scaffold binding and unbinding parameters . In particular , we randomly sample each of those parameters over a range of several orders of magnitude , consistent with experimental measurements [13] . For two-site ( de ) phosphorylation systems with scaffold , our numerical simulations suggest ( Figure 3A ) : ( 1 ) for every single bistable system found , the rate of phosphorylation from to , , is larger than the rate of dephosphorylation per unit phosphatase from to ( and from to ) , ; ( 2 ) is also almost always larger than , the rate of phosphorylation from to ; ( 3 ) the scaffold dissociation constant is low for ( ) and high for ( ) ; ( 4 ) the total substrate concentration is larger than the total scaffold concentration . In summary , the above features in parameters indicate a fast flow , together with a small flow out of , ensuring that there is an accumulation of phosphorylated protein in the scaffold-unbound state ( Figure 3B ) . Similarly , the scaffold-bound , unphosphorylated protein accumulates due to a low scaffold dissociation rate . This configuration can give rise to multiple steady states , where in fact most of the protein accumulates at either or ( data not shown ) . It is remarkable that this is the only common conformation giving rise to bistability under the chosen parameter regime . Due to the mass conservation of total substrate , a conformation in which all variables are present in either high or low concentrations is precluded . The fact that we do not allow for phosphorylation off scaffold also breaks some of the possible symmetries . It is also significant that bistability is rarely observed when . In fact , for large relative amounts of , it holds that and the system becomes approximately linear . Hence in the limit it cannot have two discrete stable steady states . It is worth pointing out that although the cell membrane was considered a suitable scaffold for ultrasensitive behavior in [16] , it may not itself be a good scaffold for bistable behavior , since must be limiting for bistability . Given that the cell membrane has a relatively large surface area , it is not likely that binding sites on the membrane will be saturated by a given membrane-binding regulatory protein . Thus , under the hypotheses of this model , employing the plasma membrane as a scaffold would be unlikely to aid in the promotion of bistability . On the other hand , the scaffold may well be a membrane-bound protein available in limited concentration . This effectively recruits the substrate onto the membrane while limiting the total amount of scaffold . We now consider the full mass action ( MA ) models and to what extent the addition of a scaffold facilitates bistable behavior . First we show that , at least for some sets of parameters , a scaffold allows a monostable multisite system to become bistable . By examining the dose-response curve as a function of the total kinase concentration , no bistability is observed for the system MA-NS ( Figure 4A ) . However , in the presence of the scaffold , with the same phosphorylation parameters ( except now phosphorylation takes place only on the scaffold ) , the response curve presents bistability for a range of values of ( Figure 4B ) . We randomize every parameter in the system over several orders of magnitude ( see the Methods section for full ranges ) , in order to find whether this behavior is typical . One preliminary result was that for no bistable behavior was found computationally for either MA-NS or MA-S , consistent with the theoretical findings for LR-NS and LR-S in the section on monostable topologies . Therefore in the following we focus on systems with multiple sites . In Figure 4C , we compare the behavior of the simplified system LR-S to that of MA-S , and we find that the dose response of both systems becomes very similar for large values of the Michaelis constant . This parameter is essential in the quantitative study of enzymes and constitutes the substrate concentration at which the enzymatic reaction takes place at half the maximal rate . It is important to note that most enzymes have been experimentally found to have a between and ( Section 8 . 4 in [31] ) . When is relatively large in a given enzymatic reaction , the flow rate from the substrate to the product can be estimated to be [32] , i . e . the detailed mass action model MA becomes similar to the linear rate model LR ( both in the presence and in the absence of a scaffold ) . See also a more detailed mathematical analysis in Section 3 of Text S1 . We are particularly interested in bistable behavior with a significant distance between steady states . To this end , we define and as the highest and lowest stable steady state values of in the case that multiple steady states exist . We restrict our definition of bistability to the case where . This definition is biologically relevant . Imagine a biological circuit with two stable steady states that are close to each other . This system is likely to have similar properties as one with a single stable steady state . It is well known that bistability can give rise to cell differentiation or other types of cellular decision-making . The underlying premise is that one of the proteins in the system , say , the most phosphorylated version of the substrate , is responsible for activating a downstream response that triggers one of the two possible cellular behaviors . If this protein is not present in sufficient concentration , the other cellular behavior should result . Therefore in practice , bistability by itself is not enough , but the two different steady states ( or at least the key active proteins ) should be sufficiently different from each other ( Figure 4D ) . In order to systematically compare different systems , we classify the models according to the value of the different enzymatic reactions . Thus enzymatic parameters are chosen randomly in such a way that all the individual values lie within a specified range of one order of magnitude . For , the system tends to be bistable even without the addition of a scaffold , and adding a scaffold decreases the probability for bistability ( Table 1 ) . However , for , the likelihood of bistability in the scaffold model ( 13 . 2% for n = 5 ) is several times that of the model without a scaffold ( 2 . 2% ) . For , the effect of adding a scaffold becomes much more pronounced . Simulations based on a set of 500 randomly chosen parameters for each entry in Table 1 indicate that in the absence of a scaffold the system is monostable for such . We next increase the number of randomly chosen parameters to 100000 for the range of , without scaffold and no bistability is found . Remarkably , if a scaffold is considered in the same circumstances the probability of bistability leaps up to 18 . 4% for n = 5 , which is significant considering that the on and off rates as well as the total protein concentrations are randomly varied over several orders of magnitude . If the phosphatase acts only off the scaffold , the probability for bistability further increases to 29 . 8% . Similar results as in Table 1 are found when the assumption of a sufficient ratio between the steady states is dropped , see Table S3 . Also , analogous results were found when the off-scaffold phosphorylation rate is low but nonzero as well as when all lie within ranges of two orders of magnitude ( data not shown ) . It should be noted that the value of is often important only with respect to the concentration of the corresponding substrate . Here we have assumed ranges for from 1 to 10 ( see the Methods section ) , and it is possible that a relevant measure for the results in the table is . In Table 2 , we repeat the same analysis as in Table 1 but classifying the parameter sets by this ratio instead of . We find that whenever , that is when , there is no bistability without scaffold , but the addition of a scaffold does allow a significant likelihood for bistability . Notice also that these results hold regardless of the dimensionality of parameter space or of the geometry of the set of bistable parameters , since we are merely measuring the proportion parameter sets that yield bistable systems . As a matter of reference , if the fraction of bistable parameter sets under given conditions is around 10% and 500 samples are taken , one can expect about 1 . 3% of standard deviation between the sampled result and the actual fraction .
In cellular signal transduction , multiple , consecutively-acting components of a signaling pathway are often physically organized into complexes by scaffold proteins . Here , by exploring various models of multisite ( de ) phosphorylation with scaffold , we conclude that under the following specific conditions the presence of a scaffold can enhance bistability of multisite phosphorylation systems . Notice that certain relations among the various parameters are also consistently preserved . For instance , the larger phosphorylation rate for the second site suggests an allosteric behavior between the substrate and kinase . Several different models of bistability in protein networks are described in [39] . In [40] , protein sequestration is considered as a means to obtaining bistability in an apoptosis network . Another approach was carried out in [41] for the MAPK system , where the activity of MEK is inhibited by unphosphorylated ERK acting as a scaffold . These systems are similar in spirit to this work , although they likely exploit a different mechanism for bistability . For instance , in [41] bistability takes place largely because the substrate is allowed to phosphorylate the scaffold and alter its binding activity , a key feedback component that we do not assume here . Also in that model the scaffold must be in excess of the substrate for bistability ( [41] , Figure 3A ) , whereas in our system we have the opposite requirement . See also the work in [42] , where a 25-fold parameter variation analysis is carried out for a MAPK model to determine the likelihood of behaviors such as bistability and oscillations . Another important aspect to consider in these chemical reaction systems is the effect of noise and stochastic behavior . If chemical reactions are allowed to take place in a non-deterministic way , the variables in a bistable system might switch spontaneously from one steady state to another . Here the fold-change measure introduced in the Results section is again useful: if the distance between the two steady states is increased , one can expect in general that the frequency of such spontaneous events is reduced . To the extent that the addition of a scaffold increases this distance , it may reduce the effect of noise . Also , we have found , for LR-S , that bistability is in a sense characterized within a certain parameter regime . If parameters are changed due to stochastic effects , bistability will tend to be preserved as long as the parameters remain within that regime . In that sense bistability in LR-S can be described as robust with respect to parameter noise . Notice that the simulations in Table 1 suggest that zero-order ultrasensitivity isn't just a mechanism for bistability in the traditional non-scaffold system , but the only such mechanism . This is because low values ( or low ratios ) seem to be necessary for bistability . Also , the results in Figure 3 suggest that ligand binding , as opposed to phosphorylation , could provide a framework for bistability using scaffolds . Assuming that the ligand is in high concentration , a simple model of multisite ligand binding would look very much like LR-S and the same analysis would likely apply . We have concluded that adding a scaffold has a large likelihood of turning a monostable multisite system into a bistable one , for large -to-substrate ratio . The intuition behind this result can be described as follows . Recall that for large values of the MA system resembles the LR system , with and without scaffold respectively . Suppose that a parameter regime is such that the are large , and that the relationships in Figure 3B are satisfied . Then LR-S is likely to be bistable , and the corresponding system MA-S is likely bistable as well since it resembles LR-S . On the other hand , LR-NS must be monostable because it is fully linear , and MA-NS is likely monostable too since it resembles LR-NS . Therefore for such a regime MA-S is much more likely to be bistable than MA-NS . This conclusion is further justified mathematically in Section 3 of Text S1 . Scaffolds typically do not possess any enzymatic activity themselves , but facilitate signaling between their bound components . One way in which they are thought to do this is by tethering their ligands in close spatial proximity to each other [24] . Another mechanism by which scaffolds can enhance signal transmission is to induce an allosteric conformational change in a bound substrate that reveals target residues , as exemplified by yeast Ste5 ( scaffold ) unlocking Fus3 ( substrate ) for phosphorylation by Ste7 ( kinase ) [23] , and human KSR ( scaffold ) unlocking MEK ( substrate ) for phosphorylation by RAF ( kinase ) [30] . In addition to speeding up certain rates , scaffolds may also slow down the rates of other enzymatic reactions by blocking the access of certain enzymes ( e . g . , phosphatases ) to bound ligands . Regardless of the precise mechanism by which they act , scaffolds generally exhibit two key properties examined in this work: sequestration and rate partition . By sequestration , we mean that the scaffold-bound population is separated from the unbound ( e . g . , cytoplasmic ) population , essentially creating two different compartments . Of course , if reaction rates and enzyme/substrate concentrations are the same in these two compartments , the scaffold will essentially be inert . Thus , rate partition –the ability of the scaffold to speed up or slow down the rate of enzymatic reactions by one of the mechanisms described above– is also crucial for forming an effective scaffold . The mathematical model of scaffolding employed herein features these two key elements of sequestration and rate partition . Sequestration is achieved in our model by accounting for the second order mass action binding of scaffold and substrate . Rate partition is achieved by allowing different rates of substrate modification depending on whether the substrate is bound to the scaffold or not . Our simple model does not incorporate other potentially interesting features of scaffold-mediated signaling , such as combinatorial inhibition , processive on-scaffold phosphorylation , and multi-tier scaffolding ( our model just has two tiers: a single kinase and its substrate ) . For other theoretical treatments of scaffold action , the reader is referred to the following references: [25] , [26] , [43] , [44] , . There has been considerable interest in understanding how common biochemical modules and motifs can be flexibly tuned to achieve a variety of desired outcomes [48] , [49] , [50] , [51] , [52] . The work presented here can be viewed as a contribution to this theme . For instance , if bistability were a desirable ( pro-fitness ) performance objective during an evolutionary trajectory , then a viable evolutionary strategy might be either a low- multisite phosphorylation module , or a high- scaffolded multisite phosphorylation module . On the other hand , if multistability were to be avoided , then there are still multiple ways that a module might have evolved , either with or without scaffolding , so that other desirable performance objectives ( e . g . , speed , amplification , specificity , etc ) might be maximized .
Throughout the computational modeling , we used mass action kinetics to construct the systems of differential equations associated with each individual model . Given a parameter set , in order to test for bistability we first reduced the problem to a 3-variable system of equations involving , , and , generalizing the approach described in [13] for scaffold systems; see Section 2 in Text S1 for details of this reduction for each type of model . Solutions of the reduced system were then found using Newton's Method with multiple different initial conditions for the MA models , and using polynomial numerical solvers for the LR models . Even though was used as the de facto output , we verified in hundreds of independent trials that the system is only bistable if itself admits multiple stable steady states . A key aspect of the analysis is the choice of random parameter sets over several orders of magnitude . Rates of substrate binding to an enzyme or scaffold are normally in the range of to [53] . Off-rates can vary more widely depending on specificity , and they are assumed here to range from about to [54] . For simplicity , we choose all rate constants , as well as in the LR systems , between and in these respective units . Total protein concentrations , were chosen from the range to . For the MA system , was chosen from to , and was used as a variable to plot a dose response curve as in Figure 4A , B , with values ranging from to . Within this range , 50 values of were sampled logarithmically ( i . e . , ) and for each value the steady states of the system were computed to create the dose response . It was determined for LR-S in Figure 3A that bistability is not found in practice for , therefore to optimize the results in Table 1 and Table 2 we assumed . For the same reason , we restricted the ratios of the scaffold binding and unbinding parameters according to the results of Figure 3A , i . e . , . These restrictions are relatively mild considering the wide range used for each parameter . All parameters were chosen under a logarithmic distribution—that is , using a uniform distribution for their natural logarithm . For the tables , in order to ensure that all values lie within a certain range , we generated the individual rate constants as described above , and if any was outside of the range then the parameters were randomized once more until all were in the desired interval . | The modification of a protein at multiple sites can result in a number of interesting behaviors at the cellular level , such as all-or-none responses to an external input , or two different stable cellular states in otherwise identical environments . Such behaviors can aid in many different forms of cellular decision-making , e . g . , cell differentiation or cell division . In this paper , we show that bistable behavior can be greatly enhanced by the presence of a scaffold protein , which binds to the substrate protein and either relocates it or otherwise affects the action of the modifying enzymes . The scaffold protein substantially widens the range of parameters for which bistability is observed when , a key descriptor of enzymatic activity , assumes medium to large values found in a majority of enzymes . Indeed , when was greater than the concentration of the target substrate , bistability was never observed in the absence of a scaffold . In addition to extensive computational work , we also carried out a mathematical analysis of a simplified system in order to identify the conditions under which bistability is possible . We conclude that scaffold proteins can be a simple yet very useful addition to multisite protein systems when bistability is advantageous . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"systems",
"biology",
"mathematics",
"biology",
"computational",
"biology",
"nonlinear",
"dynamics",
"genetics",
"and",
"genomics"
] | 2012 | Protein Scaffolds Can Enhance the Bistability of Multisite Phosphorylation Systems |
Several important human pathogens are represented in the Corynebacterineae suborder , including Mycobacterium tuberculosis and Corynebacterium diphtheriae . These bacteria are surrounded by a multilayered cell envelope composed of a cytoplasmic membrane , a peptidoglycan ( PG ) cell wall , a second polysaccharide layer called the arabinogalactan ( AG ) , and finally an outer membrane-like layer made of mycolic acids . Several anti-tuberculosis drugs target the biogenesis of this complex envelope , but their efficacy is declining due to resistance . New therapies are therefore needed to treat diseases caused by these organisms , and a better understanding of the mechanisms of envelope assembly should aid in their discovery . To this end , we generated the first high-density library of transposon insertion mutants in the model organism C . glutamicum . Transposon-sequencing was then used to define its essential gene set and identify loci that , when inactivated , confer hypersensitivity to ethambutol ( EMB ) , a drug that targets AG biogenesis . Among the EMBs loci were genes encoding RipC and the FtsEX complex , a PG cleaving enzyme required for proper cell division and its predicted regulator , respectively . Inactivation of the conserved steAB genes ( cgp_1603–1604 ) was also found to confer EMB hypersensitivity and cell division defects . A combination of quantitative microscopy , mutational analysis , and interaction studies indicate that SteA and SteB form a complex that localizes to the cytokinetic ring to promote cell separation by RipC-FtsEX and may coordinate its PG remodeling activity with the biogenesis of other envelope layers during cell division .
The Corynebacterineae suborder of bacteria includes many significant human pathogens , including Mycobacterium tuberculosis ( Mtb ) and Corynebacterium diphtheriae [1] . These organisms are often referred to as the mycolata due to the distinct architecture of the cell envelope that surrounds them . Their mode of growth also differs substantially from other well-studied bacterial model systems such as Escherichia coli and Bacillus subtilis . Like the traditional model bacteria , members of the Corynebacterineae have a cytoplasmic membrane that is surrounded by a crosslinked polysaccharide cell wall layer made of peptidoglycan ( PG ) . However , their PG matrix is uniquely modified with a second polysaccharide polymer called arabinogalactan ( AG ) , which serves as the anchor point for long hydrocarbon chains of mycolic acids ( MA ) . Together with other free mycolate glycolipids , the anchored MAs form an outer membrane-like layer called the mycomembrane that is reminiscent of the second membrane of Gram-negative bacteria ( Fig 1A ) [2 , 3] . With respect to growth mode , bacteria in the Corynebacterineae suborder elongate by adding new cell envelope material at the cell poles rather than at dispersed sites throughout the cell cylinder like E . coli and B . subtilis ( Fig 1B ) . Although many of the enzymatic steps required for the synthesis of envelope components have been defined in the Corynebacterineae [4] , it remains unclear how the biogenesis of the different layers is coordinated during growth and cell division . The mechanism of polar growth also remains poorly defined . Enhancing our understanding of these processes will address fundamental unsolved questions in microbiology while also providing information that is relevant to the treatment of diseases caused by mycobacteria and other Corynebacterineae . Drugs that target the assembly of the envelope feature prominently in the current therapeutic regimen for M . tuberculosis infections [5] . However , the rise of resistant organisms is eroding the efficacy of these drugs such that new classes of antibiotics are needed [5] . Given its status as a proven target , gaining new insights into the biogenesis of the mycolata cell envelope will provide the knowhow necessary for discovery programs to effectively develop next-generation therapies . The non-pathogenic bacterium Corynebacterium glutamicum ( Cglu ) has been a useful model organism for mechanistic studies of mycolata envelope assembly [3 , 4] . Relative to the major mycobacterial model Mycobacterium smegmatis ( Msmeg ) , it has a faster doubling time ( 1 vs . 3 hr ) and a more compact genome ( 3 . 3 vs . 7 . 0 Mbp ) that is likely to reduce the incidence of redundancy and aid genetic analyses . Another advantage of Cglu is that it tolerates more severe defects in the cell envelope , allowing the construction of mutants defective for AG and/or MA synthesis that have provided key insights into the biogenesis pathways for these envelope layers [6–9] . Where studies of Cglu have lagged behind those of the related mycobacteria is in the area of global genetic analyses . Only one large-scale transposon mutagenesis study was performed in Cglu thus far [10] . However , the library constructed only contained on the order of 11 , 000 mapped insertion mutants . We therefore set out to generate the first high-density library of transposon insertion mutants ( >100K insertions ) in this organism to enable transposon-sequencing ( Tn-Seq ) studies [11] for the discovery of new factors involved in envelope assembly in Cglu and other members of the Corynebacterineae . Sequencing of the transposon library following growth in rich medium allowed the identification of the essential gene set of Cglu , which will provide an important resource for the community . To begin identifying novel envelope assembly proteins , the insertion profile of the library was also analyzed following growth in the presence of sub-lethal concentrations of ethambutol ( EMB ) , a drug that disrupts AG biogenesis [12 , 13] . Several genes required for normal tolerance to EMB were identified using this chemical genetic approach , including the cell division genes ripC , ftsE , and ftsX . A similar phenotype was observed for genes of unknown function , which we have designated as ste loci ( sensitive-to-ethambutol ) . Among these genes , mutants inactivated for the steA ( cgp_1603 ) and/or steB ( cgp_1604 ) genes were found to have cell separation defects similar to those of ripC and ftsEX mutants . Using a combination of quantitative microscopy , mutational analysis , and interaction studies , we discovered that SteA and SteB form a complex that localizes to the cytokinetic ring to promote cell separation by the RipC-FtsEX complex and may coordinate its PG remodeling activity with the biogenesis of other envelope layers during cell division . The success of this initial study suggests that further global genetic analyses in Cglu will provide a rich discovery platform for generating insights into the processes of cell growth and envelope biogenesis among the Corynebacterineae .
The MB001 strain of Cglu was chosen for this analysis . It is a commonly used derivative of ATCC 13032 that has been engineered to remove three defective prophage elements [14] . We used the Tn5-based transposome system for transposon mutagenesis to yield a library of approximately 200 , 000 mutant colonies . Sequencing the transposon-chromosome junctions in the library yielded 10 , 419 , 920 reads that mapped to 200 , 940 unique insertions sites throughout the genome ( Fig 2A ) . This high-density insertion map allowed the global identification of essential genes in Cglu . Such genes were identified by plotting the number of unique insertions per annotated gene relative to gene length ( Fig 2B ) . The analysis revealed a bimodal distribution with the majority of genes displaying an increasing number of insertions proportional to their length and a small subset of genes that harbored very few insertions regardless of their size ( Figs 2B and S1 ) . Genes likely to be essential for growth in rich medium were defined as those harboring one or fewer unique insertions per 127 bp . With this cutoff , the percentage of essential genes was constant across a broad-range of gene lengths except for those shorter than 100bp ( S1 Fig ) . For these smaller genes , an essentiality designation likely reflects technical factors of gene size and transposon density rather than biological function . We therefore excluded genes shorter than 100 bp from the final essential gene list along with tRNAs , rRNAs , and insertion elements . With these exclusions ( 131 in all , S1 Table ) , a total of 322 essential genes were identified in Cglu ( S1 Table ) , which corresponds to roughly 11% of the total gene content , a percentage that is in-line with similar analyses of other bacteria , including Mtb [15–18] . Of the 490 total essential genes in Mtb , 445 have homologs in Cglu . However , only roughly half ( 231/445 ) of these are essential in Cglu ( S2 Table ) . Thus , consistent with the conditional dispensability of the arabinan component of the AG and the MA layer of the envelope [6–9] , Cglu is more tolerant to mutations and therefore should broadly enable studies aimed at defining the biological function of essential Mtb genes . Support for the validity of our essential gene designations came from the fact that many genes expected to play vital roles in bacterial physiology were identified as such . For example , widely conserved genes in the division-cell wall ( dcw ) cluster with known essential roles in PG precursor biosynthesis ( murC , murD , murE , murF and mraY ) or cell division ( divIC , ftsI , ftsQ and ftsZ ) were found in to be depleted of transposon insertions in the profiling data ( Fig 2C ) . Surprisingly , the Tn-Seq data indicated that ftsW , which encodes a cell division PG polymerase [19] essential for the growth of most bacteria , had a high number of insertions mapping within the reading frame , suggesting that it may not be essential in Cglu . Upon closer inspection , all of the insertions were found to be located in the 3’ end of the gene ( Fig 2D ) . Notably , FtsW in Cglu and other Corynebacterineae possesses an extended C-terminal sequence absent from bacteria such as E . coli and B . subtilis ( S2 Fig ) . In mycobacteria , this positively charged tail was previously shown to bind FtsZ and proposed to regulate its assembly into the Z-ring [20 , 21] , the key cytoskeletal structure required for cell division [22] . Our results suggest that the FtsW-FtsZ interaction may not be essential for the viability of Cglu . Based upon the example of ftsW , we computationally searched for genes in which only 5’ end of the reading frame was essential . This analysis led to the identification of 17 “domain essential” genes ( S1 Table ) bringing the total essential gene content of Cglu to 339 ( Fig 2E ) . Ethambutol ( EMB ) inhibits the function of arabinosyl transferases responsible for the polymerization of arabinan chain of AG [12 , 13] as well as lipoarabinomannan ( LAM ) , an arabinan-containing glycolipid found in the envelope of the Corynebacterineae [23 , 24] . Recent work has found that EMB treatment inhibits polar growth in Cglu and mycobacteria and causes defects in daughter cell separation [25 , 26] . We therefore reasoned that new factors involved in growth or division might be identified as loci causing EMB hypersensitivity when inactivated . To this end , the Cglu transposon library was propagated in the presence of a sublethal concentration of EMB ( 0 . 3 μg/ml ) and then analyzed by transposon sequencing . Loci associated with sensitivity to ethambutol ( ste genes ) were then identified as those displaying a significant reduction in mapped insertions following EMB treatment relative to an untreated control . A total of 49 ste loci were identified in the Tn-Seq analysis ( Figs 3A and S3A and S3 Table ) . As an indication that the screen worked as expected , several genes encoding enzymes in the AG biogenesis pathway were identified as hits along with other loci encoding factors with predicted roles in envelope biogenesis , including two loci implicated in daughter cell separation: cgp_1735 and ftsE-ftsX . The cgp_1735 gene , which we will refer to as ripC , is homologous to a gene of the same name in mycobacteria and encodes an exported protein with an N-terminal coiled-coil domain and a C-terminal PG endopeptidase domain that cuts crosslinks in the cell wall matrix ( Fig 3A and 3B ) [27] . The ftsE and ftsX genes encode an ABC-transporter-like complex ( FtsEX ) ( Fig 3A and 3B ) that was originally implicated in the control of cell wall hydrolase activity in E . coli [28] and Streptococcus pneumoniae [29] . FtsEX was subsequently found to be involved in the activation of PG hydrolase activity in a number of organisms , including mycobacteria [30] , and was recently suggested to function in complex with RipC to promote cell separation in Cglu [31] . To validate the screening results , deletion alleles for several of the identified ste genes were constructed and the corresponding mutants were tested for EMB hypersensitivity . All of the deletion mutants constructed were found to display some degree of sensitivity to EMB ( Figs 3C , S3B ) , indicating that the screen was faithfully identifying loci responsive to EMB . The morphology of each mutant was also assessed by fluorescence microscopy following staining with the membrane dye FM4-64 . The number of septa per cell was also quantified following growth in the presence of a fluorescent D-amino acid ( FDAA ) , which labels septa and other sites of new cell wall synthesis [32] . Indicative of their expected cell separation defect [26 , 27 , 31] , mutants deleted for ripC or the ftsEX locus yielded a population of cells that were longer than average relative to wild-type and possessed one or more unresolved septum ( Fig 4A and 4B ) . As previously observed [26] , cellular compartments that were ‘trapped’ in the middle of chaining cells failed to elongate properly , presumably because their cell poles remained attached to the neighboring cells . Among the other ste mutants constructed , only deletions in steA and steB were found to cause a cell separation defect that resembled cells inactivated for ripC or ftsEX ( Fig 4A and 4B ) . The steA and steB genes are located in an apparent operon that is highly conserved within the Corynebacteriales order ( S4A Fig ) . SteA is a predicted to be a bitopic transmembrane protein with an N-terminal cytoplasmic domain and a small C-terminal domain at the membrane surface ( Fig 3B ) . SteB is also a single pass transmembrane protein , but it is predicted to have a small N-terminal tail in the cytoplasm and a soluble C-terminal domain exposed to the envelope ( Fig 3B ) . As expected from the Tn-Seq data where steA displayed a greater depletion of transposon insertions following EMB treatment than steB ( Fig 3A ) , a ΔsteA mutant was found to have a more severe EMB-sensitivity phenotype than a ΔsteB mutant . The double ΔsteAB mutant was as sensitive to EMB as a single ΔsteA mutant , indicating that steA is epistatic to steB ( S5 Fig ) . Importantly , the phenotypes of all steA/B deletion mutants could be complemented upon ectopic expression of the missing gene ( s ) , indicating that the phenotypes of the deletions were not due to adverse effects on the expression of nearby genes ( S5 Fig ) . Given the importance of SteA and SteB for cell division and their conservation among the Corynebacterineae , we chose to investigate their functions further . Studies of the other ste loci will be pursued as part of a separate line of investigation . In addition to RipC , Cglu encodes a second PG hydrolase called RipA that plays a role in cell separation [27] . Unlike a ΔripC mutant , cells inactivated for RipA alone were not found to display a separation defect [27] . However , deletion of ripA in a ΔripC background was found to exacerbate the cell separation defect displayed by cells in which only RipC was inactivated [27] . We re-created these mutants in the MB001 background and confirmed these findings ( S6 Fig ) . Furthermore , we found that although deletion of ripA alone did not change the EMB-sensitivity of MB001 cells , it enhanced the EMB sensitivity of ΔripC cells ( Fig 5 ) . These results suggest that RipA and RipC contribute to distinct cell separation pathways . To determine whether SteA and SteB contribute to one of these pathways or form yet another mode of cell separation , we combined the ste deletions with those of ripC or ripA and assessed their EMB sensitivity . Deletion of ripA was found to enhance the EMB sensitivity of cells deleted for steA or steB , but ΔsteA ΔripC and ΔsteB ΔripC mutants displayed the same EMB sensitivity as the single ste deletion mutants ( Fig 5 ) . We therefore conclude that SteA and SteB are likely functioning as part of the RipC cell separation pathway . Further support for this possibility is presented below . The multiple septa observed in the ΔsteA , ΔsteB , ΔripC , and ΔftsEX mutants could either be due to a defect in cell separation or to septal closure being slowed such that new ones form before the old septa are completed . Like other bacteria , cell division in Cglu is mediated by a cytokinetic ring apparatus organized and controlled by polymers of the tubulin-like protein FtsZ [33] . We monitored the lifetime of the Z-ring constriction process as a proxy for the rate of septal closure . A fluorescent monomeric superfolding variant of GFP ( msfGFP ) fused to FtsZ ( msfGFP-FtsZ ) was produced from an ectopic site on the chromosome in the wild-type and mutant strains , and the lifetime of the Z-ring from formation to resolution was monitored by time-lapse microscopy . In the same cells , the progress of septum completion was monitored with a fluorescent trehalose reporter , 6-TMR-Tre [34] . This labeled molecule is incorporated into the non-covalently attached trehalose glycolipids that form the outer leaflet of the mycomembrane . Previous work has shown that 6-TMR-Tre infiltrates the newly completed septum through perforations in the peripheral peptidoglycan just before the daughter cells abruptly separate via a mechanical fracturing event called V-snapping [26] . The time delay between 6-TMR-Tre infiltration and V-snapping therefore provides a useful kinetic parameter with which to characterize the efficiency of the cell separation process . In wild-type cells and all of the mutants assayed , the characteristic lifetime of the Z-ring constriction process remained unchanged ( Fig 6A and 6B ) . However , while wild-type and the ΔripA mutant showed the expected short lag between septal infiltration of trehalose glycolipid and V-snapping ( 7 . 6 and 4 . 8 min , respectively ) , the time between these two events was significantly extended in the ΔsteA , ΔsteB , ΔripC , and ΔftsEX mutants ( between 31 . 7 and 45 min ) ( Fig 6C and 6D ) . We therefore conclude that the cell division defect in these mutants is the same and related to the cell separation process , not slowed septal closure . To investigate whether SteA and SteB are likely to be playing a direct role in the cell separation process , we assessed their subcellular localization . Fusions with mScarlet-I ( mScar ) [35] were constructed and produced from the attB1 site in the chromosome . Both mScar-SteA and mScar-SteB fusions restored the normal cell length distribution to cells lacking the native protein ( S7A–S7C Fig ) . The mScar-SteA fusion was also found to reduce the number of observed septa per cell back to normal ( Fig 7A ) . However , there was a notable difference in the ability of the fusions to rescue the EMB-sensitivity of mutant cells . The mScar-SteB fusion came close to fully complementing the EMBS phenotype of ΔsteB mutants ( S7D Fig ) . By contrast , mScar-SteA only weakly rescued the EMB-sensitivity of ΔsteA cells ( Fig 7A ) , indicating that there is not a perfect correspondence between cell separation and EMB resistance . We infer from this result that SteA may have a second function unrelated to cell separation and that the fusion may be defective for this activity ( See Discussion ) . Cells expressing mScar-SteA or mScar-SteB fusions displayed one of two localization patterns . In recently separated cells that remained tethered to their siblings , the mScar signal decorated the cell periphery , consistent with a membrane localization ( Fig 7B and 7C ) . In cells that were presumably further along in the division cycle , cells displayed tight band of fluorescence at the prospective division site with minimal signal observed along the cell periphery ( Fig 7B and 7C ) . Thus , SteA and SteB are recruited to the cytokinetic ring and are likely to be directly participating in the division process . We next investigated the timing of SteA recruitment to the division site relative to other measurable cell division events: Z-ring formation assessed using msfGFP-FtsZ , septal PG construction assessed by labeling of the septum with the FDAA called NADA [32] , and the infiltration of labeled trehalose-glycolipids into cell septum . The analysis was carried out using population level measurements of individually labeled cells to determine the cell length ( a proxy for age ) at which 50% of the population displayed label localized at the division site . As expected , Z-ring formation , septal PG synthesis , and infiltration of trehalose-glycolipids at the cell septum occurred in apparent chronological order with characteristic cell lengths of septal labeling of 1 . 76 , 2 . 75 , and 3 . 49 μm , respectively ( Fig 7D ) . The characteristic cell length of mScar-SteA recruitment to the division site was 2 . 64 μm , which coincides with the inception of septal PG assembly ( Fig 7D ) . To further investigate the relative timing of these two events , we simultaneously monitored nascent PG synthesis and mScar-SteA localization at the single-cell level in a time-lapsed experiment . Cells producing mScar-SteA were imaged within a microfluidic device , which allowed constant infusion with media containing NADA . Consistent with the population level analysis , in every one of the more than thirty cells that were manually monitored in time-lapse , the mScar-SteA and NADA signals appeared at the division site simultaneously ( Fig 7E ) . Furthermore , the mScar-SteA signal was observed to disperse from the cell septum immediately following V-snapping ( Fig 7E ) , indicating that SteA does not linger at the new cell poles following division . Next , we investigated the domain requirements for SteA localization and function . A series of mScar-SteA variants were constructed in which different parts of the protein were either deleted or substituted with another domain ( Fig 7A ) . Even though all of the protein fusions accumulated to comparable levels in Cglu ( S8 Fig ) , only the full-length mScar-SteA protein was capable of rescuing the the cell separation phenotype of the ΔsteA mutant or partially rescuing EMB hypersensitivity ( Fig 7A ) . In accordance with the functionality of the fusions , only mScar-SteA showed specific enrichment at the septum ( S9 Fig ) . The N-terminal domain of SteA ( mScar-SteAN ) displayed diffuse cytoplasmic signal , whereas mScar fused to the C-terminal transmembrane domain of SteA ( mScar-TMA ) labeled the peripheral membrane as well as septal membranes ( S9 Fig ) . To test whether the native SteA transmembrane domain plays a role in SteA function or localization beyond anchoring the N-terminal domain to the membrane , it was replaced by unrelated transmembrane domains from two putative tail-anchored membrane proteins of Streptomycetes coelicolor: SecE ( TM1 ) and PkaB ( TM2 ) [36] . These constructs , mScar-SteAN-TM1 and mScar-SteAN-TM2 , displayed a peripheral localization signal consistent with membrane recruitment , but they were not functional and did not promote specific septal labeling ( Figs 7A and S9 ) . We conclude that septal localization of SteA is critical for proper cell separation and that both the N-terminal and transmembrane domains of the proteins are required for its function . To determine whether any of the other factors required for cell separation are required for the recruitment of SteA to the division site , we visualized the localization of mScar-SteA in mutants inactivated for SteB , RipC , or FtsEX . All of these mutant strains contain multiple septa resulting from their cell separation defect . In each case mScar-SteA labeled these septa ( Fig 8A ) . However , we noticed that the intensity of mScar-SteA septal bands in the ΔsteB mutant was consistently lower when compared to the other two mutants . The lower septal mScar-SteA signal in ΔsteB cells was not due to a reduction in the overall mScar-SteA protein level ( S10 Fig ) . Rather , we measured a statistically significant increase in the non-septal mScar-SteA signal in the ΔsteB mutant ( 2 . 74 ± 0 . 17 , 23 cells ) compared to those of ΔftsEX ( 1 . 88 ± 0 . 09 , 18 cells ) or ΔripC cells ( 2 . 19 ± 0 . 14 , 24 cells ) ( Fig 8B ) . As a result , the ratio of septal mScar-SteA signal to background in the ΔsteB mutant ( 1 . 07 ± 0 . 37 ) was significantly reduced compared to mutants inactivated for ftsEX ( 2 . 20 ± 0 . 72 ) or ripC ( 1 . 85 ± 0 . 72 ) ( Fig 8C ) . These results indicate that SteB is required for optimal recruitment of SteA to the division site . The observation that SteA retains some ability to localize in the absence of SteB may explain why the ΔsteB mutant exhibits a milder EMB sensitivity phenotype than cells lacking SteA . The requirement of SteB for robust recruitment of SteA to the division site suggested that the two proteins might interact at the septum . To investigate their potential interaction further , we employed a two hybrid assay we recently developed called POLAR based on PopZ-linked apical recruitment in E . coli . This method takes advantage of the polar organizing protein called PopZ from Caulobacter cresentus [37–39] and its ability to form polar foci in E . coli . Bait proteins are fused to the H3H4 peptide of PopZ and sfGFP . When they are produced in E . coli cells that also make full-length PopZ , the bait fusions are recruited to the cell pole by the H3H4 peptide-PopZ interaction ( Fig 9A ) . Prey proteins are fused to mScar and their interaction with the bait is assessed based on whether or not they are also recruited to the pole ( Fig 9A ) . The assay is thus similar to previous cytological interaction detection systems based on DivIVA fusions [40 , 41] , but it benefits form the more robust and specific polar accumulation of PopZ in E . coli relative to DivIVA . An in-depth description of the POLAR method and the types of protein-protein interactions it can detect will be presented in a separate report . To assess the SteA-SteB interaction , H3H4-sfGFP-SteA was used as the bait . The fusion successfully resulted in the recruitment of SteB-mScar to the cell pole whereas a control membrane-anchored bait protein H3H4-sfGFP-TM did not ( Fig 9B ) . Additionally , a control periplasmic mScar protein was also not recruited to the pole by H3H4-sfGFP-SteA ( Fig 9B ) . In addition to visual inspection of individual cells , the POLAR results can also be viewed at a population level using a demograph in which quantified fluorescence traces of cells are stacked according to cell length and visualized as a heat map . Such population level results also support a specific interaction between the H3H4-sfGFP-SteA bait and SteB-mScar ( Fig 9B ) . Given that E . coli is only distantly related to Cglu and unlikely to encode interaction partners for Cglu proteins , we conclude from the POLAR results that SteA and SteB are likely to interact directly to form a complex . We next wanted to assess the ability of the SteAB complex to interact with RipC . However , the assay was complicated by the RipC-mScar localization pattern in cells producing the control bait fusion H3H4-msfGFP-TM , which it is not expected to interact with . The bait displayed a primarily unipolar localization signal as expected , but RipC-mScar localized to both poles in a manner uncorrelated to the pattern of the control bait ( Fig 10A and 10B ) . Even though cells in the demograph were oriented such that the pole with higher bait accumulation was oriented to the right , RipC-mScar intensity was distributed roughly equally between the two poles ( Fig 10A ) . This bipolar localization of RipC-mScar in the E . coli periplasm likely reflects some level of aggregation due to the predicted intrinsically disordered and coiled coil domains present in the N-terminal portion of RipC . Accordingly , a RipC truncation mutant lacking the C-terminal NlpC/p60 catalytic domain ( ssDsbA-mScar-RipCN ) also showed bipolar localization when it was secreted to the E . coli periplasm ( S11A Fig ) . Nevertheless , when H3H4-msfGFP-FtsEX was used as the bait , the mScar-RipC signal became biased to the pole with the highest bait signal . In this case the correlation index ( CI ) between the FtsEX bait and the RipC prey localization pattern was 0 . 96 ± 0 . 04 , much higher that the 0 . 14 ± 0 . 12 index value between RipC and the control bait ( Fig 10B ) . Given that FtsEX and RipC are known to interact robustly in mycobacteria , we conclude that the POLAR assay can accurately detect RipC interactions despite its tendency to generate a bipolar signal when fused to mScar . We therefore assessed whether or not RipC interacts with SteAB using H3H4-msfGFP-SteA as a bait co-expressed with untagged SteB . As with FtsEX , the intensity of the polar RipC-mScar signal was highly correlated with the H3H4-msfGFP-SteA bait signal ( CI = 0 . 64 ± 0 . 07 ) , albeit to a slightly lesser degree than FtsEX ( Fig 10B ) . A similar correlation of polar signals ( CI = 0 . 67 ± 0 . 12 ) was obtained when a prey harboring only the N-terminal domain of RipC was used instead of the full-length protein ( S11A and S11B Fig ) . This result indicates that RipC interacts with the SteAB complex primarily if not entirely through its N-terminal domain . To determine which component ( s ) of the SteAB complex directly contacts RipC , we used the N-terminal domain of RipC fused to H3H4-msfGFP via a TM domain ( H3H4-msfGFP-TM-RipCN ) as the bait . It recruited SteB-mScar but not mScar-SteA prey to the cell poles ( Fig 10C ) , suggesting that SteB directly binds RipC via its N-terminal coiled-coil rich domain . This interaction requires the C-terminal part of SteB as a SteB variant lacking the extracellular domain ( SteBN-mScar ) failed to be recruited to the cell pole by the RipCN bait ( Fig 10C ) . Notably , SteBN-mScar retains its ability to interact with SteA ( S11C Fig ) . Using the POLAR assay , we were unable to detect a direct interaction between FtsEX and SteA or SteB ( S11D Fig ) . Overall , the interaction data suggest that SteAB forms part of the FtsEX-RipC complex at the septum in Cglu to promote septal PG remodeling and daughter cell separation via V-snapping ( Fig 10D ) .
A better understanding of cell envelope biogenesis in the Corynebacterineae will enable the development of novel treatments for diseases caused by Mtb , nontuberculosis mycobacteria , and pathogenic corynebacteria . Over the years , Cglu has been a useful model for elucidating essential steps in the assembly of the complex and multilayered envelope unique to this class of bacteria . To enhance our ability to use Cglu as a system for discovering fundamental mechanisms required for proper envelope assembly in the mycolata , we generated the first high-density library of transposon mutants in Cglu and performed a global genetic analysis of gene function in this organism using Tn-Seq . By defining the essential gene content of Cglu for comparison with similar datasets available for Mtb [18 , 42 , 43] , the results provide an important resource for researchers studying the growth of this group of organisms . To focus the analysis on envelope assembly , we also used Tn-Seq to identify loci that when inactivated result in hypersensitivity to the AG biogenesis inhibitor EMB . The utility of the datasets generated was demonstrated through the identification of new and conserved components of the division machinery that are required for proper daughter cell separation . Like other members of the Corynebacterineae , Cglu cells complete cell division by rapidly splitting at their division septum to form two daughters [44] . The process is aptly referred to as V-snapping , and was recently characterized in detail using time-lapse microscopy and fluorescent reporters to label different layers of the envelope [26] . It was found that the envelope layers were assembled sequentially at the division site . The PG layer was the first component of the septum to be observed , followed shortly afterwards by reporters of covalently linked mycolic acids , which presumably report on both the AG layer and its attached lipids being assembled at the division site . Finally , free labeled trehalose glycolipids were observed to infiltrate the septum from the surrounding mycomembrane , and V-snapping separation took place shortly afterwards . Diffusion of trehelose glycolipids into the septum was associated with the visualization of perforations in the peripheral PG layer . What forms these initial imperfections in the PG is not clear , but their enlargement appears to require the PG hydrolase RipC [26] . Ultimately , these perforations are thought to elicit the mechanical fracture of the septum and the rapid V-snapping event [26 , 44] . The mechanisms that control the activity of cell separation hydrolases like RipC have been investigated in diverse organisms . In many bacteria , a key role for the FtsEX ABC-transporter like complex has been uncovered . In E . coli and S . pneumoniae , FtsEX associates with EnvC and PcsB , respectively [28 , 29] . Both partner proteins have an N-terminal coiled-coil domain that associates with an external loop domain of FtsX and a C-terminal domain with homology to PG hydrolases [28 , 29] . This effector domain of PcsB is thought to directly cleave PG whereas for EnvC it has been shown to activate cell wall cleavage by amidases [45] . In both cases , the proteins localize to the division site to promote daughter cell separation , and this function requires the ATPase activity of FtsE [28 , 29] . It has therefore been proposed that FtsEX promotes cell wall remodeling using ATP-driven conformational changes in the complex to activate PG hydrolysis by its partner protein . A similar role for FtsEX in cell division has since been uncovered in C . cresentus [46] . Additionally , FtsEX has been implicated in controlling cell wall cleavage by CwlO in B . subtilis [47] , but in this case it is required to promote cell wall expansion during cell elongation rather than cleavage of the cell wall septum . In the Corynebacterineae , the external loop domain of Mtb FtsX has been shown to interact with RipC in vitro and to modestly enhance its PG cleavage activity [30] . RipC has also been shown to be recruited to the division site in Cglu and inactivation of RipC or FtsEX has been associated with a cell separation defect [27 , 31] . Here , we have identified SteA and SteB as new components of the cytokinetic ring that promote the V-snapping process . Phenotypically , mutants lacking SteA or SteB are largely indistinguishable from those inactivated for RipC or the FtsEX complex . Interaction studies using the POLAR assay indicate that SteA and SteB are likely to form a complex and that the external domain of the SteB component is also likely to interact with the N-terminal domain of RipC . Our interaction results also suggest that RipC and the FtsEX complex interact in Cglu . It is interesting to note that the N-terminal domain of RipC is significantly longer that the corresponding domains of EnvC and PcsB ( S12 Fig ) , suggesting that this extension might allow for the simultaneous interaction of RipC with SteB and FtsEX . Based on all of the observed interactions , we propose that a multiprotein complex of SteAB-RipC-FtsEX forms at the division site to promote cell wall cleavage and generate the septal imperfections that eventually lead to its mechanical fracture and V-snapping . In E . coli , FtsEX has been shown to interact with the FtsA protein , a highly conserved component of the divisome that functions in part to anchor filaments of FtsZ to the membrane [48] . The FtsEX-FtsA interaction has been implicated in the activation of PG synthesis by the cytokinetic ring [49] . Therefore , the FtsEX complex in E . coli is thought to link the processes of cell wall synthesis and remodeling at the division site [49] . Notably , Corynebacterineae lack FtsA , but almost universally encode SteAB . Thus , although the precise function of the SteAB complex in cell separation remains to be determined , an attractive possibility is that SteAB serves as an alternative connector that links RipC-FtsEX with other functions of the division machinery . A clue to this possible connection comes from the observation that an mScar-SteA fusion corrected the cell separation defect of a ΔsteA mutant but not its hypersensitivity to the AG synthesis inhibitor EMB . Thus , rather than PG synthesis , SteAB may link RipC-FtsEX activity with AG biogenesis . Intriguingly , a complex of two small membrane proteins ( SweC-SweD ) unrelated to SteAB was recently found to be required for the cell wall expansion activity of the FtsEX-CwlO complex in B . subtilis [50] . Being involved in cell elongation as opposed to cell division , this PG remodeling system is also unlikely to interface with FtsA . Thus , the use of accessory membrane protein partners to modulate the function of FtsEX complexes may be a common feature employed by bacteria when the system is not connected with FtsA in the context of the division machinery . Overall , our results with SteAB highlight the potential of employing global genetic approaches in Cglu to provide insight into the mechanisms of mycolata envelope assembly . Continued mining of this dataset as well as additional Tn-Seq screens will likely lead to a wealth of new discoveries that should help us learn more about how members of the Corynebacterineae grow and how we can best interfere with the process for antibiotic development .
All Cglu strains used are derivatives of MB001 [14] . Unless mentioned otherwise , strains were grown in Brain-Heart Infusion Medium ( BHI ) that was supplemented with 15 μg/mL kanamycin ( Kan ) , when necessary . Plasmids were maintained in either DH5a ( λpir ) or S17-1 ( λpir ) . All E . coli strains used in the reported POLAR two-hybrid experiments are derivatives of TB28 [51] and are grown in LB ( 1% tryptone , 0 . 5% yeast extract , 0 . 5% NaCl ) . Whenever necessary , antibiotics were used at 25 ( Kan ) , 15 ( chloramphenicol; Cm ) , 50 ( ampicillin; Amp ) or 5 ( tetracycline; Tet ) μg/mL . Growth conditions for microscopy experiments are described in the figure legends and the appropriate Methods section . Detailed information about plasmid constructions can be found in S1 Text . All strains are listed in S4 Table and all plasmids are listed in S5 Table . We used the temperature-sensitive plasmid pCRD206 to perform allelic replacement in Cglu essential as described previously [52] with some modifications . Briefly , sequences corresponding to approximately 700 bp upstream and downstream of the desired deletion were inserted into pCRD206 . The resulting plasmid was transformed into the appropriate recipient strain . Transformants were selected and propagated on BHI-Kan agarose plates at 25°C . A few colonies were then purified on BHI-Kan agarose and grown at 37°C for 36 hours . One or two of the resulting KanR colonies were then grown in BHI liquid medium lacking antibiotic at 25°C overnight . The lower temperature allows replication from the plasmid origin , which selects for a second recombination to remove the plasmid from the chromosome . An aliquot of the overnight culture ( 10 μL ) was then spread on a BHI agarose plate supplemented with 10% sucrose , which further selects against sacB in the vector , and the plate was then incubated overnight at 30°C . The resulting colonies were replica patched onto BHI and BHI-Kan plates to identify KanS colonies lacking plasmid . Deletion alleles in KanS isolates were finally confirmed by colony PCR . For complementation studies , genes were integrated at the attB1 site on the chromosome using the pK-PIM vector [53] . The plasmid was introduced into the recipient by conjugation . Plasmids were transferred into Cglu using the E . coli strain S17-1 ( λpir ) . Conjugation was performed according to [54] with some modifications . Briefly , overnight cultures of recipient strains grown in LB supplemented with 4% ( w/v ) glucose ( LB-Glu4 ) were diluted 1:5 in the same medium and incubated for two hours at 30°C . After adjusting the density to an OD600 of 2 . 0 , each culture was incubated in a 48 . 5°C water bath for 9 minutes . Overnight cultures of the donor strain were diluted 1:100 in LB supplemented with the appropriate antibiotics and grown at 37°C until OD600 reached 1 . 0 . When the recipient cultures cooled to room temperature , the donor and recipient cultures were mixed at 1:3 ratio , gently pelleted to collect cells , which were then resuspended into 150–200 μL LB-Glu4 medium . The cell suspension was pipetted onto a 0 . 4 μm sterile filter placed on a LB-Glu4 agarose plate and incubated . Following overnight incubation at 30°C , cells were collected from the filter and resuspended into LB and plated on BHI-Kan agar plates that were supplemented with 30 μg/mL nalidixic acid to select against E . coli donors . Plasmid integration was confirmed by colony PCR using a primer mixture consisting of ATTB1g-UP , ATTB1g-UN , ATTB2g-UP and ATTB2g-UN [53] . Using WT gDNA as the template , the PCR reaction generates two bands of 464 bp and 700 bp in size , which represent uninterrupted attB1 and attB2 sites respectively . Therefore , specific disappearance of the 464-bp band was used as a diagnostic for plasmid integration at the attB1 site . Cells of Cglu were grown to stationary phase , and 10 mL of this culture was diluted into 1 L BHIS ( BHI + 91 g/L sorbitol ) that was supplemented with 25 g glycine , 0 . 4 g isoniazid , and 0 . 1% Tween 80 and incubated shaking at 18°C . The culture was chilled on ice for 1 hour when the OD600 reached 0 . 5 ( typically in 16–18 hours ) . Cells were then pelleted at 4000 x g for 20 min , washed once with 500 ml chilled 10% glycerol , and then three additional times with 100 ml chilled 10% glycerol . The cell density was adjusted to an OD600 of 20 before use for electroporation . The Tn5 transposome was purchased from Epicentre ( now discontinued ) or prepared in-house using Tn5 transposase purified as described previously [55] . For the in-house preparation , the transposon was amplified by polymerase chain reaction ( PCR ) using 5’-monophosphorylated primers ME Plus 9–3’ primer ( CTGTCTCTTATACACATCTCAACCATCA ) and ME Plus 9–5’ primer ( CTGTCTCTTATACACATCTCAACCCTGA ) using the <KAN-2> transposon supplied by Epicentre as the template . The transposon was purified using the QIAquick PCR purification kit ( Qiagen ) and eluted with TE buffer ( 10 mM Tris pH 8 . 0 and 1 mM EDTA ) . The Tn5 transposome was reconstituted by vortexing a 1:1 mixture of 1 μM purified transposase [55] with the purified transposon ( 200 ng/μL ) and used immediately after a 30-minute incubation at room temperature . Consistent with another report [56] , we found that the Tn5 transposase prepared according to Picelli et al . 2014 loses activity quickly following purification . To generate mutant libraries , transformation was carried out by electroporating 1 μL of commercial or homemade Tn5 transposome with 100 μl of freshly prepared MB001 electrocompetent cells . Each transformation reaction typically generated between two to five thousand transformants . We pooled transformants from 50 electroporations performed over 4 days to generate our library consisting of approximately 230 , 000 total transformants . DNA libraries for Tn-Seq analysis were prepared by a modified version of a published protocol [15] . Genomic DNA was extracted from strains using the Wizard Genomic DNA Purification Kit ( Promega ) and further purified using Genomic DNA Clean & Concentrator ( Zymo ) . Genomic DNA was fragmented using a Qsonica Q800RS Sonicator for 12 minutes ( using a 15 second on and 15 second off pulse cycle ) at 20% amplitude . Fragmented DNA was purified with 1 . 8× volume Agentcourt AMPure XP beads ( Beckman Coulter , Inc . ) and eluded into 30 μl water . Purified fragmented DNA was then treated with terminal deoxynucleotidyl transferase ( TdT; Promega ) in a 20 μl reaction with 1 μL 9 . 5mM dCTP/0 . 5mM ddCTP , 4 μl 5× TdT reaction buffer and 0 . 5 μl rTdT at 37°C for 1h , then at 75°C for 20min . TdT-treated DNA was purified with Performa DTR Gel Filtration Cartridge ( EdgeBio ) . Purified , TdT-treated DNA was used as a template in a PCR reaction to amplify the transposon junctions using the Easy-A Hi-Fi Cloning System ( Agilent Technologies ) . The primers used were: PolyG-1st-1 5’-GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGGGGGGGGGGGGGGGG-3’ and Tn5-1st-1 5’-ACCTGCAGGCATGCAAGCTTCAGGG-3’ . A second nested PCR was then performed to further amplify the transposon junctions and append the sequencing barcode . The primers used were generic NEBNext Multiplex Oligos for Illumina ( NEB ) and: Tn5-2nd-1 5’-AATGATACGGCGACCACCGAGATCTACACTCTTTTCAGGGTTGAGATGTGTATAAGAGA-3’ . The final product was run on a 2% agarose gel , and fragments ranging from 200–500 bp were gel purified using QIAquick Gel Extraction Kit ( Qiagen ) . Libraries were sequenced at the Tufts University Core Facility on a HiSeq 2500 ( Illumina ) or a Miseq ( Illumina ) on a 1× 50 single end run . All sequencing data generated in this study are deposited in Sequence Read Archive under accession PRJNA548135 . Reads were mapped to the C . glutamicum MB001 genome ( NCBI NC_022040 . 1 ) . To identify EMB sensitive genes , we calculated the fold change in reads between the ethambutol conditions and the no drug conditions . We also performed a Mann-Whitney U test to determine whether differences in the insertion profiles were statistically significant . Genes in which insertions were at least 3-fold enriched or depleted and had a p-value lower than 0 . 05 by Mann-Whitney were defined as hits . Visual inspection of transposon insertion profiles was performed with the Sanger Artemis Genome Browser and Annotation tool . Each strain harbors a replicative plasmid that expresses the bait and another genome-integrated plasmid that expresses the prey . We simultaneously introduced , by electroporation , both plasmids into the TB28 recipient that carried the helper plasmid pAH69 [57] . Transformants that integrated the prey plasmid at the HK022 phage attachment site and also harbored the bait plasmid were selected for by plating on LB agarose plates that contained Cm and Tet at 37°C , which prevented the replication of pAH69 . Electrocompetent TB28/pAH69 cells were prepared by first diluting an overnight culture into LB supplemented with Amp and incubated shaking at 30°C . When OD600 reached 0 . 15 , the culture was shifted to 42°C for 30 minutes to induce integrase expression . The culture was then chilled on ice for 1 hour . Cells were then collected by centrifugation ( 4000 x g ) for 20 min , washed once with 500 ml chilled 10% glycerol , and then three additional times with 100 ml chilled 10% glycerol . Growth conditions and staining procedures prior to microscopy are described in the figure legends . Prior to imaging , cells were immobilized on 2% agarose pads containing the appropriate growth medium , and covered with #1 . 5 coverslips [58] . Images were cropped and adjusted using FIJI software [59] . Microscopy: Images were obtained using a Nikon Ti inverted microscope that is fitted with a Nikon motorized stage with an OkoLab gas incubator with a slide insert attachment , an Andor Zyla 4 . 2 Plus sCMOS camera , Lumencore SpectraX LED Illumination , Plan Apo lambda 100x/1 . 45 Oil Ph3 DM objective lens , and Nikon Elements 4 . 30 acquisition software . Images in the green and red channels were taken using Chroma 49002 and 49008 filter cubes , respectively . The microscope was maintained at 30°C using a custom-made environmental control chamber . Microfluidic time-lapsed experiment: Specific experimental setups are described in the figure legends . Cells were loaded into the CellASIC Onix B04 microfluidic plates ( Millipore Sigma ) that were attached to the microscope described above using using a multi-well insert . Each imaging chamber was flushed with the appropriate growth medium before loading the cells . During the course of the time-lapse , appropriate growth media were supplied to the cells using a constant flow rate of 60 psi . To avoid reduction of growth rate due to phototoxicity , we reduced the intensity of the excitation light using neutral density filters ( at least ND8 ) in all our experiments . Single-cell image analysis: Measurements of cell lengths and fluorescence signals at the single cell level were carried out using Oufti [60] . To identify cells with septal signals , the Oufti output was additionally analyzed using peakFinder , which is an add-on from Oufti , to identify the position of fluorescence peaks along the length of the cells . Custom-written Matlab code was used to further identify cells with a fluorescence peak that is located within 0 . 3–0 . 7 position within the normalized length of the cell . Plots of cummulative percentage of cells with a septal signal as a function of cell length were generated in MATLAB . The curves were fitted with the Hill coefficient using MATLAB as described in the figure legend . Demograph generation: Using custom-written MATLAB code , cells were arranged from top to bottom according to their cell lengths . Additionally , in demographs of both the bait and prey , each cell was oriented such that the cell pole with the higher bait intensity was located on the right . Correlation of bait and prey signal distribution between the cell poles: Only signals within 780 nm from the cell poles were used for the analysis . We first calculated the fractional difference in the bait signal between the left and right poles as designated in the demograph , using the equation dbait=BL−BR0 . 5× ( BL+BR ) , in which BL and BR represents the bait signals at the left and the right cell pole , respectively . The polar prey signals were analyzed similarly using the equation dprey=PL−PR0 . 5× ( PL+PR ) , in which PL and PR represents the bait signals at the left and the right cell pole , respectively . Finally , the correlation index , CI , was calculated using the equation CI=dpreydbait ( CI = 1: complete positive correlation , CI = 0: no correlation and CI = -1: complete negative correlation ) . Overnight cultures of cells expressing mScar fusions to SteA , SteA variants , or SteB were diluted 1:1000 in fresh BHI and incubated shaking at 30°C until the optical density reached 0 . 5 . An aliquot of cells ( 25 mL ) was harvested by centrifugation and resuspended in 100 μL Lysis Buffer A ( 20 mM Tris 7 . 5 , 10 mM EDTA , 1 mg/ml lysozyme , 1 mM PMSF , 10 μg/ml DNAseI and 100 μg/ml RNAseA ) . The cell suspensions were incubated at 37°C for 30 min before an equal volume of 2x SDS loading buffer was added . The samples were sonicated for 10 min to reduce the viscosity before they were resolved on a 4–20% Criterion TGX Precast Midi Protein Gel ( Biorad ) . The gel was washed in H2O for 10 min before it was imaged with the Cy3 and Cy5 channels on a Typhoon FLA Scanner ( GE Healthcare ) . The image was scaled and cropped using FIJI [59] . For the phylogenetic tree showing the distribution of SteA , SteB and RecA homologs , the amino acid sequences of SteA ( cgp_1603 ) , SteB ( cgp_1604 ) and RecA ( cgp_2141 ) were used as a query in a BLASTp search against the NCBI “non redundant” ( nr ) database [61] with an e-value cutoff of 1e-4 for each protein . A list of all the taxa for which significant BLAST results were found was then sorted . We used a complex and diverse set of 1773 bacterial taxa called “Representative Genomes” that is available on NCBI ( ftp://ftp . ncbi . nlm . nih . gov/blast/db/ , Representative_Genomes . 00 . tar . gz ) . The phylogenetic tree was constructed using PhyloT ( http://phylot . biobyte . de/ ) and BLASTp results were plotted against the tree . The tree was visualized and annotated using iToL ( http://itol . embl . de/ ) [62] . A database of 1542 fully assembled representative bacterial genomes was created ( https://www . ncbi . nlm . nih . gov/assembly ) . The same queries as for the phylogenetic tree were used in a tBLASTn against this database . The start and stop positions of the alignment were sorted for each genome and the distances between steA and steB or steA and recA were calculated for each organism . | The pathways involved in bacterial surface assembly are critical for cell morphogenesis and serve as attractive targets for antibiotic development . Bacteria in the suborder Corynebacterineae , which includes important pathogens like Mycobacterium tuberculosis , possess a unique multilayered surface structure . In addition to the common peptidoglycan cell wall , they have an attached polysaccharide layer called arabinogalactan and an outer membrane made of mycolic acids . To enhance our understanding of cell surface biogenesis in these bacteria , we performed a global genetic analysis of gene function in the model system Corynebacterium glutamicum ( Cglu ) using transposon sequencing . In addition to defining the essential gene set in this organism , our analysis also identified SteA and SteB as components of the cytokinetic ring . These factors are conserved among the Corynebacterineae , and our results reveal that they play a critical role in the final stages of cytokinesis by promoting remodeling of the peptidoglycan layer at the division site . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"cell",
"walls",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"light",
"microscopy",
"microscopy",
"corynebacteria",
"genetic",
"elements",
"cellular",
"structures",
"and",
"organelles",
"bacteria",
"research",
"and",
"analysis",
"methods",
"fluore... | 2019 | Identification of new components of the RipC-FtsEX cell separation pathway of Corynebacterineae |
Hepatitis B virus ( HBV ) is a common cause of liver diseases , including chronic hepatitis , steatosis , fibrosis , cirrhosis , and hepatocellular carcinoma ( HCC ) . HBV chronically infects about 240 million people worldwide , posing a major global health problem . The current standard antiviral therapy effectively inhibits HBV replication but does not eliminate the virus unlike direct-acting antivirals ( DAA ) for curing hepatitis C . Our previous studies have demonstrated that human apolipoprotein E ( apoE ) plays important roles in hepatitis C virus infection and morphogenesis . In the present study , we have found that apoE is also associated with HBV and is required for efficient HBV infection . An apoE-specific monoclonal antibody was able to capture HBV similar to anti-HBs . More importantly , apoE monoclonal antibody could effectively block HBV infection , resulting in a greater than 90% reduction of HBV infectivity . Likewise , silencing of apoE expression or knockout of apoE gene by CRISPR/Cas9 resulted in a greater than 90% reduction of HBV infection and more than 80% decrease of HBV production , which could be fully restored by ectopic apoE expression . However , apoE silencing or knockout did not significantly affect HBV DNA replication or the production of nonenveloped ( naked ) nucleocapsids . These findings demonstrate that human apoE promotes HBV infection and production . We speculate that apoE may also play a role in persistent HBV infection by evading host immune response similar to its role in the HCV life cycle and pathogenesis . Inhibitors interfering with apoE biogenesis , secretion , and/or binding to receptors may serve as antivirals for elimination of chronic HBV infection .
Hepatitis B virus ( HBV ) infection continues to pose a major global health problem despite of the availability of effective HBV vaccine and antiviral drugs consisting of interferon ( IFN ) and nucleoside analogs ( NAs ) . Currently , there are more than 240 million people chronically infected with HBV worldwide [1] . HBV vaccine has greatly reduced the number of new HBV infections and hepatocellular carcinoma ( HCC ) cases but does not offer therapeutic benefit to those chronically infected with HBV . Current antiviral regimens with NAs can effectively suppress HBV replication but are not curative unlike direct-acting antivirals ( DAAs ) for hepatitis C [2 , 3] . Individuals with chronic HBV infection are at a substantial risk for progression to cirrhosis and HCC [4] . The World health organization has called for the elimination of viral hepatitis as a public health threat by 2030 [5] . The biggest challenge in eradicating chronic HBV infection is the elimination of its covalently closed circular DNA ( cccDNA ) , which is the molecular basis for viral persistence [6 , 7] . The current standard antiviral therapies are not sufficient for a complete or functional cure of chronic hepatitis B [2] . New classes of effective and safe antiviral drugs are urgently needed in order for the clearance of HBV . It is conceivable that effective antiviral therapy for curing chronic hepatitis B will likely require a combination of several drugs targeting different viral and/or cellular factors [6] . Thus , a more thorough understanding of HBV biology and the identification of novel targets are keys to the discovery and development of more effective anti-HBV drugs . HBV belongs to the Hepadnaviridae family , a large group of small enveloped DNA viruses with a partially double-stranded DNA genome of about 3 . 2 kb [8] . Over the years , a great deal of new knowledge has been obtained about the underlying molecular mechanisms of HBV DNA replication [8] , which undergoes reverse transcription of a pregenomic RNA ( pgRNA ) intermediate [9] . It encodes its own DNA polymerase that possesses both RNA-dependent DNA polymerase and RNase H activities , similar to reverse transcriptase of retroviruses . Additionally , the viral polymerase contains a terminal protein ( TP ) domain at its N-terminus that recognizes a unique RNA structure element designated epsilon ( ε ) located in the 5’-terminal repeat region of pgRNA and acts as a protein primer to initiate the viral minus-strand DNA synthesis [10] . Upon infection , the polymerase protein is removed from the relaxed circular DNA ( rcDNA ) genome in the cytoplasm , which is subsequently transported into the nucleus and converted to cccDNA [11 , 12] . The rcDNA to cccDNA conversion is believed to be carried out by cellular enzymes , including DNA polymerase and ligase [13 , 14] , although the underlying molecular mechanism of cccDNA synthesis and maintenance remains unknown . The HBV cccDNA serves as a template for transcription of all viral mRNAs and pgRNA by the cellular Pol II polymerase . The viral mRNAs and pgRNA encode 7 proteins , three different forms ( L , M , and S ) of envelope proteins ( HBs ) , preCore ( HBe precursor ) , core ( HBc ) , polymerase ( P ) , and X protein ( HBx ) [8] . The pgRNA together with the attached viral polymerase is encapsidated by the core protein to form a nucleocapsid in which reverse transcription of the pgRNA takes place , resulting in the virion rcDNA genome [12] . The syntheses of viral RNA and DNA are modulated by many different cellular proteins [8] . In contrast to the advances made in HBV DNA replication , relatively little is known about cellular factors important for HBV infection and morphogenesis largely due to the lack of robust cell culture systems of HBV infection and propagation . The discovery of sodium taurocholate cotransporting polypeptide ( NTCP ) as the HBV receptor has been a landmark advance in HBV research in recent years [15] . The expression of NTCP in nonpermissive human and murine hepatocytes conferred HBV susceptibility [15–24] . Recently , we have developed stable NTCP-expressing HepG2 ( HepG2NTCP ) and AML12 ( immortalized mouse hepatocyte ) cell lines that support robust HBV infection , replication , and morphogenesis in the presence of dimethylsulphoxide ( DMSO ) and hydrocortisone [25] . Using this newly developed HBV cell culture system and gene-specific small interfering RNAs ( siRNAs ) , we have profiled a number of cellular genes in the modulation of HBV infection and morphogenesis . Interestingly , apoE-specific siRNAs were found to effectively reduce HBV infection similar to what we had previously observed for hepatitis C virus ( HCV ) [26] . Human apoE is a plasma exchangeable apolipoprotein associated with lipoproteins of various densities ( VLDL , LDL , and HDL ) . It is a 34 kDa ( 299 amino acids ) apoprotein containing a 22 kDa N-terminal domain ( residues 1–191 ) that is recognized by receptors and a 10 kDa C-terminal domain ( residues 222–299 ) that interacts with phospholipids [27–30] . There are three major apoE isoforms ( apoE2 , E3 , and E4 ) in humans [31–33] . The N-terminal domain consists of four α-helix bundles with the receptor-binding site located in helix 4 [34] . ApoE is well known to play a central role in the transport , metabolism , and homeostasis of cholesterol and other lipids by serving as a ligand for low density lipoprotein receptors ( LDLRs ) and heparan sulfate proteoglycans ( HSPGs ) . It is also involved in the repair response to tissue injury , cell growth and differentiation , immune regulation , and development of Alzheimer’s disease [35–37] . Interestingly , a number of independent studies have demonstrated that apoE is implicated in the life cycle and/or pathogenesis of different viruses [38] , including but not limited to herpes simplex virus [39] , human immunodeficiency virus [40] , and HCV [26 , 41 , 42] . In the case of HCV , we have previously demonstrated that apoE is required for efficient HCV infection and production [26 , 43] . Mechanistic studies revealed that apoE mediates HCV cell attachment by binding to HSPGs on the surface of hepatocytes and promotes HCV production through a specific interaction with HCV NS5A and E2 [41 , 43–49] . The importance and underlying molecular mechanisms of apoE in the promotion of HBV assembly , maturation , or egress remain unknown . In the present study , we aimed to determine the association of apoE with HBV and its importance in the HBV life cycle in cell culture . Findings derived from our study demonstrate that apoE is enriched in purified HBV as determined by immunoblot and co-immunoprecipitation ( co-IP ) experiments using apoE-specific monoclonal antibodies . The apoE-blocking monoclonal antibody efficiently neutralized HBV infectivity . More importantly , silencing of apoE expression or knockout of apoE gene resulted in a reduction of greater than 90% of HBV infection and production . However , silencing or knockout of apoE gene did not affect HBV DNA replication or production of nonenveloped nucleocapsids . These novel findings demonstrate that apoE is required for efficient HBV infection and production .
ApoE is predominantly produced by hepatocytes and is critical for lipid metabolism and homeostasis [35] . We and others had previously demonstrated that human apoE is incorporated into HCV particles and plays important roles in HCV infection and morphogenesis [26 , 41 , 43 , 45–47 , 49–53] . In present study , we sought to determine if apoE is also present in purified HBV . The cell culture grown HBV was concentrated with 10% PEG 8000 and was purified by cesium chloride gradient ultracentrifugation , followed by fractionation . The levels of apoE , HBcAg , LHBsAg , and HBV DNA in each fraction were determined by WB and qPCR , respectively . Additionally , the HBV infectivity of each fraction was determined by an HBV infection assay using HepG2NTCP-P3 cells , as described in our recent work [25] . The levels of HBcAg in the infected cells were subsequently determined by WB . As expected , most of apoE was detected in low-density fractions ( 1–3 ) containing the lipoproteins of various densities such as VLDL , LDL , and HDL ( Fig 1A ) . Interestingly , significant levels of apoE were found in fractions 4 to 8 in which HBcAg , LHBsAg , and HBV DNA were also detected ( Fig 1A and 1B ) . More importantly , the same fractions 4 to 8 contained infectious HBV with peak infectivity in fraction 7 ( Fig 1C ) . To exclude possible contamination of HBV with lipoproteins , HBV virions and subviral particles were pulled down by IP using anti-HBs monoclonal antibodies prior to gradient ultracentrifugation and fractionation . Again , apoE was detected in the affinity-purified HBV as determined by presence of both HBcAg and LHBsAg ( Fig 1D ) , demonstrating a bona fide association of apoE with HBV . There might be preS1-containing subviral particles in fractions 11 and 12 based on the higher ratios between LHBsAg and HBcAg than those in fractions 8 to 10 . ApoE was also detected in fractions 11 and 12 although at lower levels ( Fig 1D ) . Thus , it is still not clear if apoE interacts with HBV subviral particles . To further confirm the presence of apoE in purified HBV , fraction 7 with the highest level of infectious HBV was subjected to IP using HBsAg- and apoE-specific antibodies as well as a normal mouse IgG ( nmIgG ) as a control . The levels of HBV DNA extracted from the precipitated HBV were quantified by qPCR , while HBcAg in the precipitated HBV was detected by WB . Anti-HBs antibody pulled down greater than 95% of HBV , whereas the apoE mAb23 precipitated about 70% of HBV particles . However , similar levels of HBcAg were detected by WB between anti-HBs- and apoE mAb23-precipitated HBV ( bottom , Fig 2A ) . In contrast , normal mouse IgG failed to pull down any detectable HBV ( Fig 2A ) . These results suggest that apoE is incorporated onto HBV envelope . To further validate apoE on the HBV envelope , a trypsin digestion experiment was performed . Purified HBV in fraction 7 was treated with trypsin in the absence or presence of an envelope-disrupting detergent Triton X-100 . ApoE was completely digested by trypsin in the absence of Triton X-100 . The HBV capsids ( HBcAg ) were digested only after the viral envelope was disrupted by treatment with Triton X-100 . These findings suggest that apoE is on the HBV envelope and is likely involved in HBV infection . To determine if HBV-associated apoE is required for HBV infection , we carried out HBV neutralization experiments in PHHs and HepG2NTCP using an apoE-specific monoclonal antibody ( mAb23 ) , which was previously shown to potently block HCV infection [26 , 46] . PHH and HepG2NTCP were infected with HBV in the presence of 4% PEG . Varying concentrations ( 0 , 0 . 4 , 2 , and 10 μg/ml ) of apoE mAb23 were added during HBV infection , during and after HBV infection , or only after HBV infection , as indicated on the left of Fig 3A . A normal mouse IgG ( nmIgG ) was used as a negative control . The levels of HBcAg and HBV cccDNA in the HBV-infected cells and the levels of HBeAg and HBV DNA in the cell culture supernatants were quantified by WB , chemiluminescence immunoassay , and qPCR , respectively . Strikingly , apoE mAb23 added during and after HBV infection effectively neutralized HBV infectivity in a dose-dependent manner , resulting in 84% reduction of HBcAg ( Fig 3A top ) and 90% decrease of HBV cccDNA ( Fig 3B ) in the cell and greater than 90% lower levels of HBeAg ( Fig 3C ) and HBV DNA ( Fig 3D ) in the cell culture supernatants at concentrations up to 10 μg/ml . More significantly , apoE mAb23 neutralized HBV infectivity when added only during HBV infection as efficiently as its presence during and after HBV infection ( Fig 3A ) . However , apoE mAb23 did not affect the levels of HBcAg expression when added after HBV infection . HBV infection of HepG2NTCP cells was also efficiently blocked by apoE mAb23 when added during HBV infection ( Fig 3A bottom ) . These findings demonstrate that HBV-associated apoE does play an important role in HBV infection . The studies by others demonstrated that apoE expression on target cells was important for HCV infection [54 , 55] . The question arose whether apoE expression on target cells is also required for HBV infection . Initially , apoE expression in HepG2NTCP cells was silenced by specific siRNAs prior to HBV infection . A non-specific control ( NSC ) siRNA was used as a control . At 48 hours after siRNA transfection , HepG2NTCP cells were infected with HBV . At 4-d p . i . , the levels of apoE expression , HBcAg , and HBV cccDNA in the cell as well as HBV DNA in the supernatants were determined by WB and qPCR , respectively . As expected , apoE-specific siRNAs efficiently silenced apoE expression in a dose-dependent fashion , resulting in a reduction of apoE by 40% , 64% , and 78% at 2 , 10 , and 50 nM concentrations . In contrast , apoE expression was not affected by NSC siRNA ( Fig 4A ) . As a result , the levels of HBcAg ( Fig 4A ) and HBV cccDNA ( Fig 4B ) in the HBV-infected cells were all proportionally decreased by increasing concentrations of apoE siRNAs . The levels of HBV cccDNA were lowered by 20% , 47% , and 84% at 2 , 10 , and 50 nM of apoE siRNAs ( Fig 4B ) . Similarly , the levels of HBeAg and HBV DNA in the cell culture supernatants were decreased by greater than 90% at 50 nM of apoE-specific siRNAs ( Fig 4C and 4D ) . There is a close correlation between the reduction of apoE expression and the decrease of HBV infection , as determined by the levels of HBcAg , HBV cccDNA , HBeAg , and HBV DNA in the infected cells and supernatants ( Fig 4 ) . To determine the specificity of apoE in HBV infection , apoB was chosen as another control as we previously found that apoB was not required for HCV infection [46] . As expected , both apoB- and apoE-specific siRNAs effectively silenced their corresponding expression . However , only apoE but not apoB siRNAs greatly reduced HBV infection ( Fig 5 ) , demonstrating a specific requirement of apoE for HBV infection . To further validate the importance of apoE in HBV infection , we have also made stable apoE gene-knockout ( apoE-/- ) HepG2NTCP cell lines . As shown in Fig 6A , three apoE-/- cell lines contain either 5-nucleotide ( nt ) deletion ( -5 ) or 1-nt insertion ( +1 ) in the apoE open-reading frame ( ORF ) . The lack of apoE expression in knockout cell lines was confirmed by WB ( Fig 6B ) . The effects of apoE gene knockout on HBV infection were determined by HBV infection of these apoE-/- HepG2NTCP cell lines . Strikingly , apoE knockout resulted in 4- , 5- , and 7-folds reduction of HBcAg , respectively , in the HBV-infected cell lines ( Fig 6B ) . Similarly , the levels of HBV cccDNA in the HBV-infected apoE-/- cells were 90% lower than those in the parental HepG2NTCP cells ( Fig 6C ) . The levels of HBeAg ( Fig 6D ) and HBV DNA ( Fig 6E ) in the cell culture supernatants were reduced by greater than 90% . Collectively , these findings demonstrate that apoE expression on target cells is critical for efficient HBV infection . We and others have previously shown that apoE is critical for HCV assembly and production besides its importance in HCV infection [41 , 43 , 53 , 56 , 57] . To determine if apoE affects HBV production , its expression in the HBV-producing cell line HepAD38 was silenced by apoE-specific siRNAs . As expected , apoE siRNAs efficiently silenced apoE expression in a dose-dependent manner . In contrast , the levels of HBcAg ( Fig 7A ) and HBV DNA ( solid black bars , Fig 7B ) in the HBV-producing cells remained unchanged , suggesting that apoE did not affect HBV DNA replication in the cell . However , the levels of HBV DNA in the cell culture supernatants were lowered by 30% , 45% , and 70% at apoE siRNA concentrations of 2 , 10 , and 50 nM , respectively ( solid gray bars , Fig 7B ) . It has been previously shown that the HBV-infected or HBV-producing hepatocytes in vitro and in vivo also produced nonenveloped nucleocapsids ( naked capsids ) [58–61] , which might confound the effect of apoE silencing on HBV production . To dissect the effects of apoE silencing on HBV and capsid production , HBV virions and nonenveloped capsids were pulled down by IP using anti-HBs and anti-HBc antibodies , respectively . Interestingly , the silence of apoE expression in HepAD38 cells did not affect the secretion of nonenveloped capsids as suggested by similar levels of HBV DNA extracted from anti-HBc-pulled down capsids ( solid grey bars , Fig 7C ) . The levels of HBV DNA extracted from the anti-HBs-precipitated HBV were proportionally decreased by up to 80% at 50 nM concentration of apoE siRNAs ( solid black bars , Fig 7C ) . These results suggest that apoE is required for efficient production of HBV virions but not nonenveloped capsids . To further confirm the importance of apoE for HBV production , apoE gene was subjected to knockout in the HBV-producing HepAD38 cells similar to the above-described apoE-/- HepG2NTCP cells ( Fig 6 ) . A total of three apoE-/- HepAD38 cell lines were validated by DNA sequence analysis and the absence of apoE expression as determined by WB , including 83-nt , 85-nt , and 110-nt deletions in the apoE ORF ( Fig 8A and 8B ) . The production of HBV virions and nonenveloped capsids in the supernatants was determined by qPCR quantification of HBV DNA extracted from HBV and nonenveloped capsid , which were pulled down by IP using anti-HBs and anti-HBc antibodies , respectively . HBV production was lowered by 80–90% in the apoE-/- HepAD38 cells compared to that in the parental cells ( solid black bars , Fig 8C ) . However , the levels of HBV DNA extracted from nonenveloped capsids remain unchanged ( solid grey bars , Fig 8C ) . Taken together , these findings demonstrate that apoE is also critical for envelopment and/or production of HBV virions but does not play a role in the secretion of nonenveloped HBV capsids . The defective HBV infection and production in the apoE-/- HepG2NTCP and HepAD38 cells could be caused by potential off-target effects associated with CRISPR/Cas9 gene-editing system . To exclude this possibility , we sought to determine whether the defect of HBV infection and production could be restored by ectopic expression of apoE . The apoE-/- HepG2NTCP cells were transfected with the apoE3-expressing vector pCMV-XL5-hApoE3 . At 48-h p . t . , cells were infected with HBV , followed by the detection of apoE and HBcAg . The cell culture supernatants were collected for quantification of HBV DNA . Interestingly , over-expression of apoE3 in parental HepG2NTCP cells enhanced HBV infection by greater than 300% . More significantly , ectopic expression of apoE3 fully restored HBV infection in the apoE-/- HepG2NTCP cells ( Fig 9A and 9B ) . Similarly , over-expression of apoE in parental HepAD38 cells further increased HBV production by about 60% . Ectopic expression of apoE3 in the apoE-/- HepAD38 cells restored HBV production to 85% of that in parental HepAD38 cells ( Fig 9C and 9D ) . Collectively , these results demonstrate that apoE is specifically required for efficient HBV infection and production .
In the present study , we have obtained substantial evidence demonstrating that human apoE is associated with infectious HBV and plays an important role in HBV infection . ApoE was found to be enriched in purified HBV , as demonstrated by its co-existence with HBcAg , LHBsAg , and HBV DNA in the same fractions ( Fig 1A , 1B and 1D ) . There was a close correlation between apoE and HBV infection as shown by the peak level of HBV infectivity in fractions 6 and 7 ( Fig 1C ) , which also contained the highest levels of apoE ( Fig 1A ) . The association of apoE with HBV was further supported by a specific capture of HBV with an apoE monoclonal antibody similar to anti-HBs . Most of purified HBV could be specifically pulled down by anti-apoE and anti-HBs , as determined by the levels of HBV DNA and HBcAg in the immunoprecipitated virus ( Fig 2A ) . Interestingly , the HBV-associated apoE was sensitive to trypsin digestion unlike HBcAg which only became sensitive to trypsin digestion upon disruption of the viral envelope by treatment with triton X-100 ( Fig 2B ) . These findings suggest that apoE is likely exposed on the HBV envelope and may play a role in HBV infection . Indeed , HBV infectivity in PHHs and HepG2NTCP could be efficiently neutralized by an apoE monoclonal antibody , which was added during HBV infection or during and after HBV infection , resulting in a greater than 90% reduction of HBcAg and HBV cccDNA in the cell and HBeAg and HBV DNA in the cell culture supernatants ( Fig 3 ) . However , HBV infection was not significantly affected when the apoE-blocking antibody was added to cell culture media after HBV infection ( Fig 3A ) . These findings demonstrate that apoE on the viral envelope does play a critical role in HBV infection . Apart from the importance of the virus-associated apoE in HBV infection , apoE expression from target cells is also required for efficient HBV infection . Both silencing of apoE expression and knockout of apoE gene greatly decreased the permissiveness of HepG2NTCP cells to HBV infection ( Figs 4 and 5 ) . However , silencing of apoB expression did not affect HBV infection ( Fig 5 ) , suggesting a specific requirement of apoE for HBV infection . We have not examined the effect of apoE silencing or knockout on HBV infection in PHH or other human hepatocyte cell lines besides HepG2NTCP cell lines available in the lab . More importantly , the defect of HBV infection in the apoE-/- HepG2NTCP cells could be completely restored by ectopic expression of apoE ( Fig 9A and 9B ) . However , HBV protein expression and DNA replication were not significantly affected in apoE knockdown or knockout HepAD38 cells ( Figs 7 and 8 ) , demonstrating that apoE does not play a role in HBV protein expression or DNA replication . These findings are similar to those observed for HCV . ApoE gene knockout was also found to greatly impair the susceptibility of Huh7 . 5 cells to HCV infection but not replication [54 , 55] . Thus , it is likely that apoE may mediate HBV infection via a similar mechanism to HCV infection , which is warranted for future investigation . The underlying molecular mechanism for the apoE-mediated promotion of HBV infection is not clear , particularly for apoE expressed in target cells . In an analogy to its role in HCV infection [45 , 47 , 62] , apoE may also mediate HBV cell attachment . This possibility is supported by circumstantial evidence derived from previous studies by others . Earlier studies found that heparin was able to block HBV infection in cell culture . Similarly , cells treated with heparinase became less susceptible to HBV infection , suggesting a role of HSPG in the apoE-mediated HBV cell attachment [63 , 64] . In fact , one of the HSPG core proteins , glypican 5 ( GPC5 ) , was found as an HBV entry factor [65] . Taken together , these earlier findings suggest that HSPGs serve as HBV attachment receptors . HSPGs are known to be apoE-binding receptors and play a pivot role in the metabolism of apoE-containing lipoproteins [66–68] . Our previous studies demonstrated that apoE mediates HCV cell attachment through interaction with HSPGs [44 , 45 , 47 , 48] . However , there is no good assay to confirm the role of apoE in HBV cell attachment unlike HCV . Neither is clear how apoE expressed in target cells facilitate HBV and HCV infection . In the case of HCV , several studies by others suggested that apoE binds HCV envelope protein E1 and E2 [49 , 69] . The interaction between E1/E2 and apoE may stabilize HCV attachment mediated through the interaction between apoE and HSPGs . Similarly , our preliminary results appeared to suggest an interaction between apoE and the preS1 domain of LHBsAg . The interaction between apoE and preS1 will further promote HBV attachment mediated through the binding of apoE to the cell surface HSPGs . Alternatively , transfer of the exchangeable apoE from target cells to virions during HBV infection could also enhance HBV infection , as demonstrated by extracellular apoE-mediated promotion of HCV infection [54 , 55] . Besides its importance in HBV infection , apoE is also critical for efficient HBV production similar to its role in HCV assembly , maturation , and release [41 , 43 , 49 , 53 , 56 , 57] . Silencing of apoE expression in the HBV-producing HepAD38 cells significantly reduced HBV production in proportion to the siRNA-induced down-regulation of apoE expression ( Fig 7 ) . Likewise , apoE gene knockout decreased HBV production by 90% , as determined by the levels of HBV DNA extracted from anti-HBs-precipitated HBV virions in the cell culture supernatants among three different HepAD38 cell lines ( Fig 8 ) . However , the levels of HBcAg and HBV DNA in the apoE-silencing ( Fig 7 ) or knockout ( Fig 8 ) HepAD38 cells remained the same , demonstrating that apoE does not play a significant role in HBV protein expression or DNA replication . Neither was the secretion of nonenveloped capsids affected by silencing of apoE expression ( Fig 7C ) and knockout of apoE gene ( Fig 8C ) in the HBV-producing HepAD38 cells . Whether apoE play any role in the secretion of HBV subviral particles remains unknown . We used anti-HBs to pull down HBV virions and subviral particles but could not discriminate them by an immunoblot assay with a preS1-specific monoclonal antibody upon density gradient ultracentrifugation and fractionation ( Fig 1D ) . Nevertheless , we speculate that apoE may play a role in HBV post-envelopment during virion assembly and maturation steps similar to its function in HCV morphogenesis [52] . We and others had previously found that apoE is required for the formation of infectious HCV particles through a specific interaction with NS5A [41 , 43] . Mutations introduced to apoE and NS5A that disrupted the apoE-NS5A interaction resulted in an impairment of HCV production [41 , 43] . The studies by others suggested an interaction of apoE with HCV envelope proteins E1/E2 [49 , 69] . It is possible that the interaction between apoE and preS1 domain of the LHBsAg might promote HBV envelopment and/or egress through an unknown mechanism . Whether HBV morphogenesis and egress depend on the lipoprotein secretory pathway remains unknown . The role of VLDL secretion pathway in HCV morphogenesis and production has been controversial [46 , 70 , 71] . The siRNA-mediated knockdown of apoB expression and specific inhibitors of microsomal triglyceride transfer protein ( MTTP ) , which is essential for assembly and secretion of VLDL [72 , 73] , did not affect HCV production , as demonstrated by our previous study [46] . It will be interesting to determine if lipoproteins of different densities play any role in HBV infection and production in vivo . Whether apoE isoforms ( E2 , E3 , and E4 ) exhibit difference in HBV infection and production is warranted for further investigation . Nevertheless , it remains enigmatic how apoE promotes HBV production . The identification of the cellular gene ( s ) and/or pathway responsible for the incorporation of apoE into HBV particles may provide novel molecular target ( s ) for intervention of HBV infection . The close association of HBV with apoE or apoE-containing lipoproteins may be not only important for HBV infection and production but also viral pathogenesis by evading host immune response [74] . It was previously shown that apoE incorporated into HCV could interfere with the activity of HCV-neutralizing monoclonal antibodies in cell culture , suggesting a role of apoE in persistent HCV infection in vivo [75 , 76] . It is conceivable that the HBV-associated apoE may also play a critical role in chronic HBV infection by blocking the accessibility of HBV-neutralizing antibodies and evading cellular immune response targeting HBV-infected cells . Inhibitors that interfere with apoE biogenesis , secretion , and its interaction with cell surface receptors may serve as potential anti-HBV agents . It will be interesting to determine whether the levels of apoE and/or apoE-containing lipoproteins in the plasma of hepatitis B patients correlate with the viral load and/or the outcomes of HBV infection .
A puromycin-resistant HepG2NTCP cell line ( HepG2NTCP-P3 ) was described in our previous work [25] . The HBV-producing cell line HepAD38 was described previously [77] . Both HepG2NTCP and HepAD38 were grown in DME/F12 medium containing 10% of fetal bovine serum ( FBS , Atlanta Biologicals ) . Primary human hepatocytes ( PHHs , lot#4405C ) were purchased from Lonza and were cultured in Power Primary HEP medium ( Takara , San Francisco , CA ) . HEK293T cells were from ATCC and grown in DMEM containing 10% of FBS as previously reported [78] . Cell culture flasks and plates were coated with 50 μg/mL of rat tail collagen type I ( Corning ) . HBV obtained from HepAD38 cells were concentrated by precipitation with 10% polyethylene glycol ( PEG ) 8 , 000 ( Hampton Research ) . The genome copy numbers of HBV DNA were quantified by a real-time PCR method . ApoE-specific monoclonal antibodies 23 ( mAb23 ) and WuE4 ( ATCC ) were described in our previous work [26 , 45] . ApoE mAb23 was purified by GenScript ( Piscataway , NJ ) . HBV core-specific monoclonal antibody ( T2221 ) was from Tokyo Future Style [25] . HBV preS1 antibody ( SC57761 ) and normal mouse and rabbit IgGs were purchased from Santa Cruz Biotechnology . Human β-actin monoclonal antibody ( AC15 ) and protein G agarose were obtained from Sigma-Aldrich . HRP-conjugated goat anti-mouse antibody was from Cell Signaling . Anti-HBs antibody ( hepatitis B immune globulin , HBIG ) was from Nanyue Biopharmaceuticals Inc , China . Clarity Max Western blotting ECL substrate was purchased from Bio-Rad . ApoB- and ApoE-specific Smartpool siRNAs and a nonspecific control ( NSC ) siRNA were from Dharmacon [26] . One-Step qRT-PCR kits were from Thermo Scientific . Taq 5× master mix was from New England Biolab . HBeAg chemiluminescence immunoassay ( CLIA ) kits were purchased from Autobio Diagnostics Co . ( Zhengzhou , China ) as previously described [19 , 25] . Co-IP kits were purchased from Pierce ( Thermo-Fisher ) . Protein A magnetic beads ( Dynabeads ) were from Thermo Fisher . TRIzol reagent was from Invitrogen . Genome DNA isolation kit was from QIAGEN . Exonucleases ( ( Exo I ) and III ( Exo III ) were purchased from New England Biolabs . HBV grown from HepAD38 cells was concentrated by precipitation with 10% polyethylene glycol 8000 ( PEG 8000 ) . Concentrated HBV was used for IP of HBV and subviral particles using anti-HBs antibodies . A total of 100 μg of anti-HBs was incubated with 50 mg of protein A magnetic Dynabeads in 500 μl volume at 4°C overnight . Upon extensive washing , anti-HBs-conjugated beads were mixed with 1 ml of PEG-concentrated HBV with rotation at 4°C overnight , followed by washing with PBS three times . HBV virions and subviral particles pulled down by IP were released into 500 μl of glycine ( 50 mM ) buffer at room temperature for 10 minutes , followed by neutralization with the addition of 500 μl of Tris-HCl ( pH7 . 5 ) . Immunoprecipitated HBV or PEG-concentrated virus was subjected to purification by Cesium chloride ( CsCl ) gradient ultracentrifugation . Briefly , equal volume of 2 X CsCl solutions ( 49 . 74g CsCl dissolved to a total of 100 ml Tris-Sodium Chloride-EDTA with 0 . 05% β-Mercaptoethanol ) was mixed with equal volume of concentrated HBV , followed by centrifugation in a Beckman SW55Ti rotor at 50 , 000 rpm at 4°C for 94 h . The gradient was fractionated from bottom to top . HBcAg , LHBsAg , and apoE in each fraction were detected by Western blot ( WB ) . HBV DNA in each fraction was extracted with a QIAGEN viral DNA isolation kit and was quantified by a real-time PCR method . HBV infectivity in each fraction was determined by an HBV infection assay using HepG2NTCP-P3 cells in 24-well cell culture plates . Each fraction of 25 μL was diluted with DME/F12 medium into a 300 μL volume containing 4% PEG-8000 and was used to infect HepG2NTCP-P3 cells at 37°C for 12 hours . At 4 days post-infection ( p . i . ) , the HBV-infected HepG2NTCP-P3 cells were lysed with a RIPA buffer . Cell lysates were used for quantification of HBcAg by WB using the HBc monoclonal antibody T2221 [25] . Purified HBV particles were treated with trypsin in the absence or presence of 1% Triton X-100 for 1h at 37°C , as described previously [46] . Trypsin reactions were terminated by the addition of 1×SDS protein loading buffer . After trypsin digestion , HBcAg and apoE were determined by WB using specific monoclonal antibodies . A total of 50 μg of purified apoE mAb23 , anti-HBs , or a normal mouse IgG ( nmIgG ) was coupled to protein G agarose at room temperature for 2 hours . The antibody-conjugated agarose was then mixed with 10 μL of purified HBV ( Fraction 7 ) at 4°C overnight , similar to HCV IP [46] . After extensive washing , the precipitated HBV was used for detection of HBcAg or extraction of HBV DNA , which was subsequently quantified by qPCR using HBV sequence-specific primers and probe as described previously [13 , 15 , 25] . For pull-down of HBV virions and subviral particles as well as nonenveloped capsids by IP , 5 mg of protein A magnetic beads ( Dynabeads ) were conjugated with 10 μg of anti-HBs and anti-HBc , respectively , and then incubated with 250 μl of cell culture supernatants from apoE-silencing or apoE-/- HepAD38 cells . Upon washing with ice-cold PBS , HBV virions and naked capsids were eluted out by incubation with 100 μl of 50 mM glycine buffer , followed by neutralization with 100 μl of 1M Tris buffer ( pH7 . 5 ) . HBV DNA was extracted using QIAGEN genome DNA extraction kit . HepG2NTCP-P3 or PHH cells were seeded into 24-well cell culture plates at a density 1×105 per well at one day prior to infection . HepG2NTCP-P3 cells were infected with HBV at a multiplicity of infection ( m . o . i . ) of about 100 copies of genome equivalent in the present of 4% PEG 8000 for 12 hours except otherwise indicated . HBV-infected cells were cultured in DME/F12 medium containing 4% FBS , 1% DMSO , and 5 μg/mL hydrocortisone for 4 days or otherwise as indicated . HBV DNA in the cell culture supernatants was extracted with QIAGEN DNA isolation kits . HBV cccDNA in the cell was isolated using the Hirt method [79] , followed by treatment with exonucleases I ( Exo I ) and III ( Exo III ) , which removes DNAs with open 5’ and 3’ ends , including HBV rcDNA [13 , 25 , 80] . Chemiluminescence immunoassay kits purchased from Autobio Diagnostics Co . ( Zhengzhou , China ) were used to quantify the levels of HBeAg in the cell culture supernatants as described previously [19 , 25] . Bio-Rad protein assay dye was used to measure protein concentration of cell lysates . A total of 25 μg protein from each sample was separated by electrophoresis in SDS-PAGE and then transferred onto a polyvinylidene difluoride ( PVDF ) membrane using a semidry blotter ( Bio-Rad ) . Immunoblot analysis was done using Clarity Max western ECL substrate ( Bio-Rad ) and mouse monoclonal antibodies specific to HBcAg , LHBsAg , apoB , apoE , or β-actin . For neutralization of HBV infectivity during HBV infection , concentrated HBV in 4% PEG was used to dilute the apoE-blocking monoclonal antibody 23 ( mAb23 ) into varying concentrations ( 0 , 0 . 4 , 2 , and 10 μg/mL ) . HBV in the presence or absence of the apoE mAb23 was subsequently used to infect HepG2NTCP-P3 or PHH cells in 24-well cell culture plates . A normal mouse IgG was used as a negative control . After 12-h incubation at 37°C , the HBV-infected cells were cultured in DME/F12 medium containing 4% FBS , 1%DMSO , 5 μg/μL hydrocortisone . Additionally , apoE mAb23 was added to cell culture media after the above HBV infection in the presence of apoE mAb23 , referring as mAb23 presence during and after HBV infection . ApoE mAb23 was also tested for its inhibitory effect on HBV infection when added to cell culture media only after HBV infection . At 4-days post-infection , the levels of HBcAg in the infected cells were determined by WB . HBV cccDNA in the cell was extracted with the Hirt method . The levels of HBeAg and HBV DNA in the supernatants were quantified by chemiluminescence immunoassay and qPCR , respectively . The apoE- , apoB-specific siRNA and a NSC siRNA were described previously [26 , 46] . HepG2NTCP-P3 cells in 24-well cell culture plates were transfected with increasing concentrations ( 0 , 2 , 10 and 50 nM ) of apoE , apoB or NSC siRNA using RNAiMax reagent ( Invitrogen ) . At 48 h post-transfection ( p . t . ) , HepG2NTCP-P3 cells were infected with HBV in 4% PEG at a m . o . i . of about 100 copies of genome equivalent for 12 h at 37°C . After an additional 4-days incubation , cells were lysed for detection of apoE and HBcAg by WB , whereas the media were collected for quantifying the levels of HBV DNA and HBeAg by quantitative PCR ( qPCR ) , and chemiluminescence immunoassay , respectively . The lentiCRISPRv2-blasticidin vector previously made in the lab was digested with BsmBI . The apoE gene-specific sgRNA was designed based on the website http://crispr . mit . edu ( Feng Zhang ) . The apoE-specific sgRNA cDNA was constructed by annealing two synthetic oligonucleotides , apoE/F ( 5’-CACCGGCTTTTGGGATTACCTGCGC-3’ ) and apoE/R ( 5’-AAACCGCGCAGGTAATCCCAAAAGCC-3’ ) . The double-stranded DNA fragment was inserted into the BsmBI-digested lentiCRISPRv2-blasticidin vector in the same way as described previously [44] . Resulting plasmid DNA construct was confirmed by DNA sequence analysis and was designated pLentiCRISPRv2-Blast/ApoE-sgRNA . For lentiviral production , HEK293T cells in 6-well plates were transfected with pVSVg , psPAX2 , and pLentiCRISPRv2-Blast/ApoE-sgRNA using lipofectamine 3000 reagent according to the manufacturer’s instructions ( Invitrogen ) . At 72 h p . t . , the supernatant was collected and filtered through a 0 . 45-m low-protein-binding membrane unit ( Millipore ) . Resulting lentivirus expressing apoE sgRNA and CRISPR/Cas9 was used to transduce HepAD38 and HepG2NTCP cells . Stable apoE-knockout cell lines were selected with 5 μg/mL blasticidin and were confirmed by WB and DNA sequence analyses . An apoE expression plasmid pCMV-XL5-apoE3 was made previously [43] . The parental and apoE-/- HepG2NTCP cells were seeded at 1 × 105 cells/well in 24-well cell culture plates . A total of 1 μg of pCMV-XL5-apoE3 was transfected into HepG2NTCPcells using Lipofectamine 3000 reagent . At 48 h p . t . , the DNA-transfected cells were infected with HBV in 4% PEG 8000 at about 100 copies of genome equivalent for 12 h at 37°C . After incubation for 4 days , cells were lysed for measuring apoE and HBc by WB , whereas the cell culture supernatants were collected for quantifying the levels of HBV DNA by qPCR , respectively . HBV cccDNA was extracted by the Hirt method [79] . HBV DNA was quantified by a real-time PCR method using two HBV-specific primers: 5′-GAGTGTGGATTCGCACTCC-3′ ( forward ) and 5′-GAGG CGAGGGAGTTCTTCT-3′ ( reverse ) . HBV cccDNA was quantified using primers 5′- TCATCTGCCGGACCGTGTGC-3′ ( forward ) and 5′- TCCCGATACAGAGC TGAGGCGG-3′ ( reverse ) and probe HBV-cccP: 5’-FAM-TTCAAGCCTCCAAG CTGTGCC TTGGGTGG C-TAMRA -3’ . Internal control Mitochondrial DNA was quantified using primer mitoF: 5’-CCCTCTCGGCCCTCCTAATAACCT-3′ ( forward ) and mitoR: 5’-GCCTTCTCGTATAACATC GCGTCA-3’ ( reverse ) . The qPCR was carried out with TaqMan SYBR Green Master Mix ( Applied Biosystems ) or iTaq Universal Probes Supermix ( Bio-Rad ) and at 95 °C for 10 min ( 1 cycle ) , and 95 °C for 15 s and 60 °C for 60 s ( 40 cycles ) . Statistical analyses were conducted using Prism5 software ( Graphpad Software ) . Results are shown as means ± standard deviations ( SD ) of the data obtained from three independent experiments or triplicates as indicated . Comparisons between samples were done using the paired two-tailed t test . P values of < 0 . 01 were considered statistically very significant . | Little is known about the importance of host factors in HBV infection , assembly , and release due to the lack of robust cell culture models of HBV infection and propagation . The discovery of sodium taurocholate cotransporting polypeptide ( NTCP ) as the HBV receptor has made it possible to determine the roles of cellular genes in the modulation of HBV infection , assembly , maturation , and release . Through characterization of purified HBV , we found that human apoE is enriched in HBV and is incorporated onto the virus envelope , as demonstrated by immunoblot and immunoprecipitation using apoE-specific monoclonal antibodies and trypsin digestion . More importantly , HBV infection could be efficiently blocked by an apoE-specific monoclonal antibody or by silencing apoE expression and apoE gene knockout in the cell . These findings suggest that apoE likely mediates HBV cell attachment similar to its role in hepatitis C virus infection as previously demonstrated by us and others . Besides its importance in HBV infection , apoE is also required for efficient HBV production . Down-regulation of apoE expression or knockout of apoE gene from the HBV-producing hepatocytes severely impaired HBV production . Collectively , our findings demonstrate for the first time that apoE promotes both HBV infection and production . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"pathogens",
"biological",
"cultures",
"immunology",
"microbiology",
"viral",
"structure",
"hepatitis",
"b",
"virus",
"viruses",
"cell",
... | 2019 | Human apolipoprotein E promotes hepatitis B virus infection and production |
Nerve cells produce electrical impulses ( “spikes” ) through the coordinated opening and closing of ion channels . Markov processes with voltage-dependent transition rates capture the stochasticity of spike generation at the cost of complex , time-consuming simulations . Schmandt and Galán introduced a novel method , based on the stochastic shielding approximation , as a fast , accurate method for generating approximate sample paths with excellent first and second moment agreement to exact stochastic simulations . We previously analyzed the mathematical basis for the method’s remarkable accuracy , and showed that for models with a Gaussian noise approximation , the stationary variance of the occupancy at each vertex in the ion channel state graph could be written as a sum of distinct contributions from each edge in the graph . We extend this analysis to arbitrary discrete population models with first-order kinetics . The resulting decomposition allows us to rank the “importance” of each edge’s contribution to the variance of the current under stationary conditions . In most cases , transitions between open ( conducting ) and closed ( non-conducting ) states make the greatest contributions to the variance , but there are exceptions . In a 5-state model of the nicotinic acetylcholine receptor , at low agonist concentration , a pair of “hidden” transitions ( between two closed states ) makes a greater contribution to the variance than any of the open-closed transitions . We exhaustively investigate this “edge importance reversal” phenomenon in simplified 3-state models , and obtain an exact formula for the contribution of each edge to the variance of the open state . Two conditions contribute to reversals: the opening rate should be faster than all other rates in the system , and the closed state leading to the opening rate should be sparsely occupied . When edge importance reversal occurs , current fluctuations are dominated by a slow noise component arising from the hidden transitions .
Variability in dynamical biological systems is ubiquitous . Discrete state , continuous time Markov process models are used throughout cell biology , neuroscience , and ecology to represent the random dynamics of processes transitioning among multiple locations or states [1–3] . Examples include transitions between states defined by degree of phosphorylation and subcellular compartment location in a signaling network [4] , transitions among several conducting and non-conducting states in populations of ion channels [5] , random genetic drift across a fitness landscape [6] , random dispersal of mobile populations [7] , and many other processes [8] . Often fluctuations arise at the molecular level , whether from discrete population effects , thermal ( Brownian ) effects , or deterministic high dimensional nonlinear dynamics ( chaos ) at microscopic scales . In general , nonlinear stochastic systems cannot be solved mathematically in closed form . Even if we limit ourselves to Markov processes , i . e . models for which the probability distribution of future states is independent of the past history , given the current state ( meaning that the current state is as complete a description of the process as possible , and no additional “hidden” variables exist ) , the effects of noise on biological dynamics must usually be studied via computer simulation . However , exhaustively simulating all noise sources within a given molecular level Markov process is often computationally prohibitive . Hence there is a need for complexity reduction methods . In this paper we investigate a complexity reduction method for discrete state , continuous time Markov process models known as stochastic shielding which we summarize in the next paragraph [9 , 10] . Complexity reduction for such models aims to capture the essential dynamics and stochastic properties of a system via a simpler representation , with minimal loss of accuracy . There is substantial literature on the approximation of complex random walk models with simpler models by mapping states of the full model to the nodes of a smaller set of states [11–23] . This includes coarse-graining of complex networks [11–13] , elimination of fast variables via quasi-steady state approximation [24] , marginalization of a partially observed Markov process through the solution of a filtering problem [25] , the k-core decomposition ( first proposed in [14] and shown to be effective for visualization in [15] ) , and various clustering algorithms that have been developed recently [16–20] ( reviewed by [21] ) . Aggregation of tightly interconnected nodes and adiabatic elimination of fast variables lead to reduced models that are no longer Markovian [20 , 22] . As another approach , one may eliminate rarely visited nodes , again leading to a reduction in the number of states [23] . Stochastic shielding provides an alternative approach by simplifying the description of the noise driving the process , while preserving the Markov property , by removing from the model those fluctuations that are not directly observable [9] . As illustrated in Fig 1 , rather than reduce the number of nodes in the graph , the stochastic shielding approximation reduces the number of independent noise sources used to drive the stochastic process on the graph , while preserving the dynamical behavior of a particular projection of the random process . As discussed in more detail in Methods §Summary of Stochastic Shielding in the Langevin Case , in the Langevin approximation for a time homogeneous first-order transition network , the population fraction occupying states 1 , … , n is a vector X ( t ) ∈ R n satisfying d X d t = L X + ∑ k ∈ E B k ξ k ( 1 ) where L , the graph Laplacian , captures the mean flux along each directed edge k ∈ E ( edge set ) . The matrix Bk gives the effects of fluctuations ξk around the mean flux along the kth edge . The noise terms , ξk , are independent , white and Gaussian , one for each directed edge . Given an observable of interest , represented by a vector M ∈ R n , the stochastic shielding approximation consists in finding a partition of the edge set into edges of primary importance ( E 1 ) and secondary importance ( E 2 ) that gives an approximate process Y ( t ) ∈ R n satisfying d Y d t = L Y + ∑ k ∈ E 1 B k ξ k , Y ( 0 ) = X ( 0 ) ( 2 ) by neglecting the noise forcing along the edges of secondary importance . Such an approximation typically creates a ( small ) pathwise discrepency relative to X that can be quantified by our edge importance measure , also defined in Methods and discussed in more detail below [10] . The stochastic shielding approximation exploits filtering properties intrinsic to any network . Given an observable defined on the network ( for example the indicator function for a subset of states representing nodes of interest ) , the fluctuations in population flux along some edges will have a greater impact on fluctuations in the observable , while other edges’ fluctuations will have a lesser impact . Hence the network “shields” the observable from some fluctuations , which may therefore be ignored with little loss of accuracy . To put it another way , the effects of a fluctuation in the movements of populations far removed from a location of interest do not directly affect the fluctuations in the population of interest; their effect reaches the observed nodes only via the indirect effect of influencing the population immediately surrounding the node or nodes of interest . One may view the source of fluctuations ( relative to the average flux along a given edge ) as independent noise forcing associated with each edge in the graph [10] . Edges that connect nodes that are indistinguishable , with respect to the measurement vector M , are themselves not directly observable . The fluctuations in rates of transition along these hidden edges are “averaged over” and their effect on the observed value ( M⊺ X ( t ) ) is reduced . This filtering effect leads to the possibility of a novel approximation scheme . Rather than approximating a random process on a graph by aggregating together subsets of nodes , we may replace the fluxes along a subset of edges with the mean flux along the respective edge . If a graph has K directed edges , there are 2K − 1 such “approximations” , as the independent noise along each edge can be either included or excluded from the approximation . Including all noise terms gives the original model , whereas excluding all noise terms gives a model with no fluctuations . Which of these 2K − 1 different approximations is the “best approximation” ? The stochastic shielding method provides the following rule: suppress the noise along those edges connecting indistinguishable nodes . We extended this method by introducing an edge importance measure that quantifies the effect of suppressing noise along each edge separately . For a linearized Langevin equation ( multidimensional Ornstein-Uhlenbeck ( OU ) process ) approximating the full population process , we showed that when the process satisfies detailed balance , the variance of the observable states can be decomposed into a sum of fluctuations attributable to each pair of directed edges in the graph . Thus , the edge importance measure allows one to rank the edges such that the most important edge contributes the most to the stationary variance of the observable states . We previously applied the stochastic shielding method to Markov processes arising in neuroscience ( Hodgkin and Huxley’s sodium and potassium ion channel models ) and processes on Erdos-Renyi random graphs [10] . However , these processes do not include significant timescale separation . In the present paper we study processes with nonuniform stationary probabilities and multiple timescales , including ion channel models with “bursty” dynamics . Separation of timescales is an important property of many neural systems [26] . For instance , many ion channels exhibit bouts of repeated channel opening and closing , interspersed by long periods of channel closure—often referred to as bursty conductances . The nicotinic acetylcholine receptor ( nAChR ) is a well studied ligand-gated ion channel that can exhibit bursty behavior [27–29] . Acetylcholine ( ACh ) is a neurotransmitter that plays a key role in motor function via this ion channel , and the opening of the nAChR channel pore requires the binding of ACh . For low acetylcholine concentration ( [ACh] ) , the nAChR is a classic example of a bursty ion channel . In the next section , we explore the robustness of the stochastic shielding phenomenon and the accuracy of the approximation under conditions of timescale separation and sparsity in the stationary distribution , by way of the edge importance measure described in [10] . We show that typical edge importance hierarchy is robust to the introduction of timescale separation for a class of simple networks , but that it can break down for more complex systems with three or more distinct timescales , such as the nAChR described above . Nevertheless we also establish that the edge importance measure remains a valid tool for analysis for arbitrary networks regardless of multiple timescales .
The shielding phenomenon leads the fluctuations associated with directly observable transitions to dominate the variance of the observable states in many networks , but this rule does not hold universally . The edge importance measure ( see Eq 42 in Methods ) provides an exact means to evaluate the applicability of stochastic shielding to any model ( Markovian , with first-order transitions ) by quantifying the effect of suppressing noise along each edge separately . This measure considers the pathwise mean square error between two trajectories: the full stochastic process with all fluctuations included , and an approximate process with a subset of fluctuations excluded . We use this measure to rank the edges in order of importance with respect to the stationary variance of the observable states . Moreover , we show that the stationary variance decomposes into a sum of contributions from each edge . This decomposition is unique and follows from a straightforward calculation that we describe and prove in Theorem 1 in the last subsection of Results . We apply the stochastic shielding method and compute the edge importance measure for the acetylcholine receptor model introduced above and for a set of simple networks ( 3-state chains ) with timescale separation . The nicotinic acetylcholine receptor is a ligand-gated ion channel and the opening of the channel pore requires the binding of acetylcholine . For low acetylcholine concentration , the nAChR is a classic example of a bursty ion channel . This channel has been described many times in the literature , and we will follow the formulation from Colquhoun and Hawkes [30] . Following Figure 4 . 1 in their paper , the channel has five states with ten possible transitions between states . The states form a graph with vertices i ∈ V = { 1 , 2 , 3 , 4 , 5 } and edges k ∈ E = { 1 , ⋯ , 10 } ( see Tables 1 and 2 ) . Fig 2A shows the transition state diagram . The channel can be bound to zero , one , or two ACh molecules . When singly or doubly bound the channel may be open or closed , whereas the unbound state is always closed . Table 1 gives the definition of the states and labels each state as open ( observable ) or closed ( unobservable ) . State 5 ( T ) is the unbound state ( closed ) , state 4 ( AT ) is singly bound ( with 1 molecule of ACh ) and closed , state 3 ( A2T ) is doubly bound and closed , state 2 ( A2R ) is doubly bound and open , and state 1 ( AR ) is singly bound and open . The measurement vector M specifies which states are open and which are closed by labeling each state with a 1 or 0 , respectively . In this case , M is given by M = ( 1 1 0 0 0 ) ⊺ ( 3 ) meaning that states 1 and 2 are open/conducting states and states 3 , 4 , and 5 are closed/non-conducting states . Table 2 gives the definition of the edges and the transition rates . Note that the ten transitions are numbered starting with the pair of transitions connecting states 1 ( AR ) and 2 ( A2R ) and moving clockwise back to state 1; these are reactions 1-8 . The last pair of transitions ( 9 and 10 ) connect states 4 ( AT ) and 5 ( T ) . We will write the per capita transition rate for the kth reaction , with source node i and destination node j , either with a single index denoting the reaction ( αk ) or with a double index denoting the source followed by the destination ( αij ) . Thus , α1 and α21 are synonymous . Burstiness is defined by the observation of isolated single channels opening and closing in bouts [29–31] . Fig 2D shows a sample trace of our model simulation exhibiting burstiness of the channel for low agonist concentration ( [ACh] = 0 . 5μ Mol ) . ( For details on the model simulation , see Numerical Methods , in Methods . ) Fig 2E zooms in on the burst in panel D labeled by the red arrow . The distribution of closed intervals shows a mixture of slow and fast timescales , requiring combinations of two or more exponentials with widely separated time constants . These time constants are related to the eigenvalues of the graph Laplacians ( see Eq 4 , and see Methods for details ) . The ratio of eigenvalues will be used as a measure of timescale separation . Fig 2B shows the presence of timescale separation at low [ACh] concentrations by plotting the ratios of the eigenvalues {λ2/λj}j = 3 , 4 , 5 . Significant timescale separation occurs when λ2/λj << 1 , or in words , when the two eigenvalues differ by at least one order of magnitude . The graph Laplacian has leading eigenvalue λ1 = 0 . For the acetylcholine receptor , and for the systems we study here , the remaining eigenvalues are real and negative , and are ordered so that 0 > λ 2 ≥ λ 3 ≥ … ≥ λ | V | , where | V | is the number of states . We apply the stochastic shielding method to the nAChR model and show that it works well for high acetylcholine concentration , but not in the bursty regime characterized by low ACh concentration . In fact , we see a reversal of edge importance at low agonist levels ( see Fig 2C and discussion below ) . In light of the network filtering effect underlying stochastic shielding , we might naïvely expect that the edges connecting states 2 and 3 , and states 1 and 4 , should contribute the most to the stationary variance of the observable states ( 1 and 2 ) , but this is not the case . There is even a regime where the observable edge pair ( edges 3 and 4 ) is only the third most important edge , as defined by our edge importance measure . Computing the edge importance measure ( Eq 42 in Methods ) , the fraction of the stationary variance contributed by edge k , requires the graph Laplacian L ( and its corresponding eigenvalues and eigenvectors ) , the noise coefficient matrix B ( defined below ) , the stationary mean flux Jk , and the measurement vector M . The graph Laplacian L as a function of ACh concentration c is L = ( - ( a 1 + k + 2 * c ) 2 k - 2 * 0 b 1 0 k + 2 * c - ( a 2 + 2 k - 2 * ) b 2 0 0 0 a 2 - ( b 2 + 2 k - 2 ) k + 2 c 0 a 1 0 2 k - 2 - ( b 1 + k + 2 c + k - 1 ) 2 k + 1 c 0 0 0 k - 1 - 2 k + 1 c ) ( 4 ) and matrix B is B = ( J 1 ζ 1 J 2 ζ 2 … J 10 ζ 10 ) , ( 5 ) where Jk = Ntot αij πi ( k ) is the stationary flux across edge k for a total population of Ntot ion channels , αij is the appropriate transition rate of reaction k ( Table 2 ) and ζk is the stoichoimetry vector for reaction k . The kth stoichoimetry vector describes how an individual moves from node i to node j in reaction k . For instance , the first two stoichoimetry vectors are ζ 1 = ( 1 − 1 0 0 0 ) ⊺ ( 6 ) ζ 2 = ( − 1 1 0 0 0 ) ⊺ , ( 7 ) which correspond to transition 1 ( an individual moves from state 2 to state 1 ) and transition 2 ( an individual moves from state 1 to state 2 ) , respectively , in Fig 2A . Note that ζ1 = −ζ2 , and this relationship holds for each edge pair in the ACh transition graph . The matrix B depends on the equilibrium population distribution π → = ( π 1 , … , π 5 ) ⊺ . Since π → is the leading eigenvector of the graph Laplacian L , the equilibrium fraction πi of the population in state i will change as a function of c ( ACh concentration ) . Lastly , recall that the measurement vector M = ( 1 1 0 0 0 ) ⊺ as described in Table 1 . Fig 2C plots the relative edge importance Rk ( fraction of the stationary variance contributed by edge k ) for each edge k ∈ {1 , … , 10} as a function of acetylcholine concentration over the range [ACh] ∈ [10−1 , 102] μMol . At high concentrations , the most important edges are those connecting the doubly bound closed state to the doubly bound open state ( edges 3 and 4 ) , that is , the edges along which transitions are directly observable . This situation is consistent with results for Hodgkin-Huxley ion channels and generic Erdos-Renyi random graphs with randomly assigned binary measurement vector [9 , 10] . In contrast , the most important edges at low concentrations are those connecting the singly bound state to the doubly bound closed state ( edges 5 and 6 in Fig 2A ) . Although transitions along this edge are not directly observable , they make a greater contribution to the stationary variance of the open state than the opening/closing transitions . Moreover , we find that edges 5 and 6 have the highest relative importance for low and intermediate concentrations , followed by edges 3 , 4 and 9 , 10 . Just below a concentration of 10 μMol , the relative importance switches so that edges 3 and 4 become the most important for higher concentrations ( ≥ 10 μMol ) . To begin to understand why the edge importance ranking changes for low [ACh] , we note that the relative importance depends heavily on state occupancy probability . As has been previously observed , one of the nodes in the 5-state nAChR model has very low occupancy probability across all agonist concentrations [32] . In particular , states 2 ( A2R ) and 5 ( T ) are the most likely states to be occupied over the range of [ACh] considered . However , state 1 ( AR , one of the open states ) has very low occupancy probability and hence is rarely visited by the process . As a result , the most likely path between the unbound/closed state 5 ( T ) and the doubly bound/open state 2 ( A2R ) is 5 → 4 → 3 → 2 . This means that transitions 7 , 8 and 1 , 2 do not happen very often . The stochastic shielding method predicts that these reactions should be important , but if they rarely happen , they contribute little to the stationary variance . Thus , their relative importance as computed by our edge importance measure is very small . Indeed , for all values of [ACh] , the equilibrium occupancy probability of state 1 , π1 is ≪ 1 . The variance of the open states for a population of Ntot channels at equilibrium is V [ Open ] = N tot ( π 1 ( 1 - π 1 ) + π 2 ( 1 - π 2 ) - 2 π 1 π 2 ) ≈ N tot π 1 ( 1 - 2 π 2 ) + N tot π 2 ( 1 - π 2 ) + O ( π 1 2 ) , as π 1 → 0 + . Although the goal of the stochastic shielding approximation is not to change the network topology by eliminating nodes as other authors have suggested [23 , 32 , 33] , when edges are “unimportant” it is natural to consider eliminating them . If all the edges to a node are unimportant , eliminating them would eliminate the node , and in this case the change in stationary variance of the open states would be approximately Ntotπ1 ( 1 − π1 ) − 2Ntotπ1π2 , if π1 is small . ( Compare to [32] , “Scheme 1” . ) The edge importance measure Rk ( for each edge k ) provides an intrinsic idea of how many edges could be suppressed in an approximation ( whether by suppressing the fluctuations generated by that edge , which is the focus here , or by removing the edge entirely ) . For the typical operating range of the nAChR , roughly 1-10 μM [ACh] , there are three transition pairs with similar edge importance ( edge pairs 3 , 4 , 5 , 6 , and 9 , 10 ) , suggesting that accurate simulations of stochastic effects would require keeping the fluctuations generated by all three of these edge pairs . The acetylcholine receptor example suggests that the inversion of edge importance is related to timescale separation . In the next subsection , we investigate the edge importance measure in the presence of timescale separation , as well as a combination of sparsely and abundantly populated vertices . We show that edge importance ranking is preserved despite the introduction of arbitrary timescale separation in simple graphs ( 3-state chains ) with per capita transition rates at two distinct timescales . As we will see , a system needs at least three distinct timescales in order to see the method break down . Nevertheless , the edge importance measure remains exact , and informative , for arbitrary networks , and can be used to extend the original stochastic shielding method to systems with timescale separation and bursty behavior . Motivated by the example of the acetylcholine receptor , we systematically study the effects of introducing timescale separation into the simplest nontrivial model to which stochastic shielding applies: the 3-state chain with one observable state ( or one pair of observable transitions into and out of the observable state ) . Specifically , we consider a discrete state , continuous time Markov jump process N ( t ) ∈ N 3 with Ntot random walkers moving independently on a graph with three nodes . See Fig 3 for an illustration of the graph , and see Methods for general notation and see S1 Supporting Information for a detailed description of the 3-state model . Here we assume that state 3 ( black disk ) is the observable state , which yields the following measurement vector: M = ( 0 0 1 ) ⊺ . If we think of this model as a simplified ion channel with three states , then the observable state is the open or conducting state of the system , and all other states are closed or non-conducting . There are four directed edges in the graph , and edge k represents a transition from source node i ( k ) to destination node j ( k ) which happens at rate αk ( or αij , see Methods for details on notation ) . We focus on the observed process M⊺N ( t ) which describes the evolution of the open state , and approximate processes that suppress noise along a subset of the four edges . In particular , we use the following two approximate processes to illustrate how stochastic shielding “usually” works: ( i ) suppress noise along edge pair 1 , 2 ( and preserve noise along edge pair 3 , 4 ) and ( ii ) suppress noise along edge pair 3 , 4 ( and preserve noise along edge pair 1 , 2 ) . In most cases ( i ) is the best approximation; we investigate here whether or not this heuristic holds universally . The mechanism of stochastic shielding can be readily understood by considering the power spectrum of the observed process M⊺N ( t ) . The relationship between the power spectrum and the covariance matrix of a stochastic process is well known; the power spectrum is the Fourier transform of its covariance [34] . The stationary covariance C of a discrete state Markov process ( such as N described above ) is given by Gadgil , et al . [8] , and satisfies the Lyapunov Eq 46 ( see Methods ) . The stationary variance R of the full and approximate observed processes has the following connection to the power spectrum: integrating over the power spectral density ( PSD ) S ( ω ) gives the stationary variance . Moreover , since the stationary variance decomposes into a sum of contributions from each edge in the graph ( R = ∑kRk where Rk is the edge importance measure of edge k given in Eq 42 ) , the power spectrum decomposes as well ( S ( ω ) = ∑k Sk ( ω ) , see Eq 66 ) . We provide more details on how the power spectrum is obtained in Methods §Numerical Methods . Fig 4B shows sample trajectories for the full process ( denoted by X , black trace ) and the two approximations ( i ) and ( ii ) described above ( denoted by X3 , 4 ( red trace ) and X1 , 2 ( blue trace ) , respectively ) in the Gaussian ( OUP ) version of the model . Fig 4A shows the corresponding power spectral contributions for the three processes: S ( ω ) is the total PSD ( shown in black ) , S3 , 4 ( ω ) is the PSD for approximation X3 , 4 ( red ) , and S1 , 2 ( ω ) is the PSD for approximation X1 , 2 ( blue ) . See Methods §Numerical Methods for details on model simulation and calculation of the power spectra . At all frequencies , the power from the observable edge pair 3 , 4 predominates , as shown by the red dashed line ( S3 , 4 ( ω ) ) closely following the black line ( total PSD ) . This spectral decomposition agrees with our edge ranking based on edge importance ( i . e . edge pair 3 , 4 contributes the most to the stationary variance ) , and illustrates why the stochastic shielding method says that the best approximation of the full process is to preserve the noise along edge pair 3 , 4 and to suppress the noise along edge pair 1 , 2 . Fig 4B illustrates the consequence in the time domain: the red trajectory closely follows the black trajectory , but the blue trajectory only captures a rough approximation of the full process . However , this situation breaks down and leads to edge importance reversal for certain bursty systems , which we aim to understand in the rest of the paper . In the remainder of this section , we show that edge importance inversion cannot be obtained by taking a 3-state chain and accelerating or decelerating any single edge , pair , or trio of edges with a single parameter ( i . e . by introducing two distinct timescales ) . As we shall see , in order to invert the edge importance as we did in the nAChR example for low agonist concentration , we need to introduce a third timescale . This will be addressed in §Generalized 3-State Model with Timescale Separation . For completeness , we may consider the same 3-state chain as in Fig 3 , except that we set the middle state ( state 2 ) to be the open/conducting state instead of state 3 . The measurement vector in this case is M = ( 0 1 0 ) ⊺ . See the left column of Fig 6 for an illustration , and note that there are five possible cases to consider . In this version of the 3-state chain , all transitions are observable since each edge connects the conducting state to a closed state , and hence , all edges should be important in terms of stochastic shielding . State 2 no longer acts as a “shield” as it did when state 3 was the conducting state . We expect that the most important edges will either depend on the parameter α or all edges will be equally important in terms of the edge importance measure . We repeat the same analysis as in the previous section and the results are shown in Fig 6 . Fig 6 has the same three column format as Fig 5 . The left column shows the 3-state diagram with accelerated/decelerated transition rates ( 1 or α as outlined in Table 3 ) where again α ∈ [10−4 , 104] . The middle column shows timescale separation as defined by the ratio of the two non-zero eigenvalues ( λ2/λ3 ) versus α . The right column shows the relative edge importance Rk versus α . In contrast to the previous cases with state 3 conducting , now we see edge importance reversal or convergence in every case . This is what we expect , given that the stochastic shielding method says that all edges are important in this version of the model . However , we find edge importance reversal in Case 10 without corresponding timescale separation since λ2 and λ3 differ by less than one order of magnitude in that case . We showed above that the presence of two distinct timescales was not sufficient to see an inversion of the edge importance in a 3-state network . However , as we show next , a network exhibiting three separate timescales can lead to edge importance reversal . In order to find examples of inversion , we study an ensemble of 3-state chains with observable state 3 ( see Fig 2 ) with arbitrary transition rates {α12 , α21 , α23 , α32} . We randomly draw the transition rates αij independently from a lognormal distribution with a given width w , that is , log ( αij ) is Gaussian distributed with mean zero and standard deviation w . Then we calculate the edge importance for each realization of transition rates for this general 3-state model and look at the instances for which R12 = R21 > R23 = R32 . Note that Rij refers to the importance measure for the edges connecting node i to node j . For an ensemble of 105 samples with log ( α i j ) ∼ N ( 0 , 10 ) ( i . e . w = 10 ) , we find that inversion of the edge importance occurs approximately 9 . 8% of the time . This observation raises a number of questions . Which factors contribute to inversion of the usual edge importance relation ( e . g . timescale separation ) ? Given an arbitrary set of transition rates , is there a canonical transformation leading to edge importance reversal ? Can we obtain an exact expression for the relative contribution of the hidden edges to the stationary variance ? The balance of this section addresses these questions . Fig 7 illustrates the distribution , for this ensemble , of several factors that might be expected to play a role in inverting edge importance . Each panel plots the relative importance of the hidden edges η = R 12 R 12 + R 23 ( 9 ) versus factors representing node occupancy , timescale separation , flux distribution , and local timescale difference . Inversion of edge importance occurs when R12 > R23 , that is , when η > 1/2 . Node Occupancy: The left column of Fig 7 plots η versus the stationary occupancy probability of each state: π3 for state 3 ( panel A ) , π2 for state 2 ( panel C ) , and the ratio π2/π3 ( panel E ) . Panel A suggests that edge importance can be inverted for any values of π3 ( mutatis mutandis π1 ) , but panel C suggests that inversion requires π2 ≲ 1/6 . Moreover , panel E indicates that inversion requires π2 < π3 ( equivalently , α23 > α32 since π2/π3 ≡ α32/α23 ) . Together , these conditions suggest that sparse occupancy of the hidden state directly connected to the observable state ( relative to the occupancy of the observable state ) contributes to inversion of edge importance . However , this condition alone is not sufficient , as shown by Example A below ( see Examples subsection ) , for which the relative importance due to the hidden edges is η = 0 . 4132 < 0 . 5 . We can extract several strict inequalities relating η to properties of the 3-state process . Maximizing η with π2 fixed , we find η ≤ ( 1 − π 2 1 + π 2 ) 2 . ( 10 ) Fig 7C shows this inequality is tight ( dashed red curve superimposed on the dots matches the upper boundary ) . In panel E , maximizing η with π2/π3 fixed , we observe that η ≤ 1 - ( π 2 / π 3 1 + π 2 / π 3 ) = π 3 π 3 + π 2 = α 23 α 23 + α 32 ( 11 ) ( dashed red curve matching boundary ) , which shows that inversion ( η > 0 . 5 ) is only possible if π2 < π3 , or equivalently , α23 > α32 . More extreme edge importance inversion requires a more extreme likelihood difference between the observable state and its neighbor or between the transition rates connecting these states . Timescale Separation: We introduce two different notions of timescale separation . First , we define ν = λ 3 / λ 2 ( 12 ) which is the ratio of the two non-zero eigenvalues of the graph Laplacian L . This quantity is shown in Fig 7B where η is plotted versus ν . ( Note ν is the reciprocal of the ratio used to define timescale separation in the previous 3-state model sections with two distinct timescales and discussed in Figs 5 and 6 ) . Large timescale separation , defined via the eigenvalues of the graph Laplacian , occurs when ν ≫ 1 . Specifically , Fig 7B shows that edge importance reversal requires timescale separation such that |λ3| ≳ 15|λ2| or ( ν ≳ 15 ) . Second , we consider the relaxation time τ i j = ( α i j + α j i ) - 1 ( 13 ) for an isolated 2-state Markov processes with rates αij , αji between the nodes i , j . The ratio of two such local relaxation times gives an alternative measure of timescale separation within the network . Specifically , consider the two possible 2-state processes in our 3-state model ( nodes 1-2 and nodes 2-3 ) . In the first system ( between nodes 1 and 2 ) , the eigenvalues of the graph Laplacian are 0 and α12 + α21 = 1/τ12 . Likewise , looking at states 2 and 3 as a 2-state Markov process yields eigenvalues 0 and α23 + α32 = 1/τ23 . Fig 7F shows the dependence of η on the ratio of the non-zero eigenvalues for these two 2-state systems . Empirically , we see that η ≤ τ 12 / τ 23 1 + τ 12 / τ 23 = τ 12 τ 12 + τ 23 ( 14 ) ( dashed red curve in Fig 7F where η is plotted versus τ12/τ23 ) . That is , inversion of the edge importance ( η > 0 . 5 ) occurs only when equilibration along the hidden edges is slower than along the observable edges ( τ12 > τ23 ) . Stationary Flux Distribution: Recall that the stationary flux along edge k is given by Jk = Ntotαkπi ( k ) . We can also represent this term as Jij , the stationary flux from node i ( k ) to node j ( k ) ( see §Notation in Methods ) . Here we define Δ J = J 12 - J 23 J 12 + J 23 ( 15 ) which is the relative fraction of the stationary flux generated by the hidden edges . In Fig 7D , we observe that the upper boundary is given by η ≤ 1 2 - Δ J 2 ( 16 ) which says that edge importance reversal ( η > 0 . 5 ) requires larger mean flux along the observable edges than along the hidden edges . In other words , the system needs to satisfy ΔJ < 0 or J12 < J23 . Reproducing edge importance reversal in 3-state chain models is advantageous because such simple Markov models can be analyzed more completely than models with greater numbers of states [23] . Fortunately , explicit expressions may be derived for the eigenvalues and eigenvectors of the 3-state chain model which allows direct calculation of η , the fraction of the stationary variance generated by the hidden edges ( see S1 Supporting Information §Explicit calculation of η for detailed derivation ) : η ≡ R 12 R 12 + R 23 = ( α 21 α 12 + α 21 ) ( α 23 α 12 + α 21 + α 23 + α 32 ) = ( π 1 π 1 + π 2 ) ( α 23 Tr [ - L ] ) ( 22 ) where Tr[L] ≡ ∑i Lii is the trace of L . The fraction in Eq 22 is a product of two factors ( denoted by F1 and F2 and shown in Fig 7G for the ensemble ) . The first factor F1 is the ratio of the speed of transition from hidden state 2 to hidden state 1 ( α21 ) to the sum of the transition rates between states 1 and 2 . Equivalently , this is the proportion of time spent in hidden state 1 relative to hidden state 2 . F1 approaches 1 as α12 decreases , which only occurs if condition 2 holds ( α12 ≪ α21 ) . The second factor F2 is the ratio of the opening transition rate ( α23 ) to the sum of the four rates . This factor is large if and only if the opening rate is much faster than the other rates , and this is exactly condition 1 ( α23 ≫ max{α12 , α21 , α32} ) . Together these two conditions bring about a reversal of edge importance ( η > 0 . 5 ) in this simple scenario . While the exact formula for the relative edge importance ( 22 ) applies only for the 3-state chain model considered here , we anticipate that analogous results may hold for more general Markov processes . We consider this question further in §Discussion . Additional insight into the error arising from different noise-suppressing approximations can be obtained by examining the power spectral density distributions of the true and approximate processes . Recalling Fig 4A in the case αij ≡ 1 , the power spectra for the full process with all noise sources included ( S , black curve ) and the approximate process with hidden edge flux noise suppressed ( S3 , 4 , red curve ) are very similar , with an order of magnitude less power arising from the hidden edges at all frequencies . In contrast , Fig 8A shows the power spectra for the 3-timescale model . In particular , it shows that at low frequencies , the power spectrum for the approximate process with visible edge flux noise suppressed ( S1 , 2 , blue curve ) is very similar to the PSD for the full process , but that the blue and red curves cross at an intermediate frequency ( ω ≈ 3 ) so the red curve dominates at high frequencies . The change in power spectral contributions is also reflected in model simulations ( see Numerical Methods for details on simulations ) . Fig 8B illustrates sample trajectories for the three processes described above: full process X ( black ) , approximation X3 , 4 ( red ) , and approximation X1 , 2 ( blue ) where α = 10 . Comparing this edge importance reversal case to the base case shown in Fig 4B , we see that the blue trajectory ( instead of the red one ) closely follows the black trajectory . Hence , X1 , 2 is the better approximation to the full process in this case . Thus , the edge importance reversal observed under the combined conditions α12 ≪ α21 and α23 ≫ max ( α12 , α21 , α32 ) may be understood as resulting from enhancement of the noise power contribution from the hidden edges at low frequencies , as well as the small amplitude of the full process’ power spectrum at high frequencies . We see a similar mechanism at work in the 5-state acetylcholine receptor model in the low-[ACh] regime ( where a hidden edge becomes more important than a visible edge ) as opposed to the high-[ACh] regime , in which the usual edge importance ordering is observed . Figs 9 and 10 show the power spectrum and Gaussian model trajectories in the high-[ACh] and low-[ACh] regimes , respectively . Here we have similar notation to the 3-state cases: X ( black ) is the full observed process ( Gaussian version ) with all sources of noise included and Xi , j is the approximate process that preserves noise on edge pair i , j but suppresses noise on all other edges . In particular , we focus on the red trace ( X3 , 4 , noise preserved on visible edges 3 , 4 ) and the blue trace ( X5 , 6 noise preserved on hidden edges 5 , 6 ) . The usual edge ordering via the edge importance measure for high [ACh] ranks edge pair 3 , 4 the most important , followed by edges 5 , 6 , then 9 , 10 ( the last two edge pairs have relative importance close to 0 and make the two lowest spectral contributions ) ; See Fig 2 . Fig 9A shows that for [ACh] = 100 μM , most of the power is attributable to the observable edge pair 3 , 4 , and this agrees with the edge importance ranking . Model trajectories in panel B illustrate that X3 , 4 is the best approximation of the full process X and that the other approximations at best only capture the mean behavior of the system . In the low-[ACh] case shown in Fig 10 ( [ACh] = 0 . 5 μM ) , however , we see the crossing of the top blue and red power spectral density curves at an intermediate frequency ( ω ≈ 2 ) . As in the 3-state case , this indicates a reversal of edge importance whereby now the hidden edge pair 5 , 6 contributes the most to the stationary variance of the observable process . Again , this change in spectral contributions is reflected in model trajectories shown in panel B . We see that the blue curve X5 , 6 closely follows the full process X , and is the best approximation in this case , but the blue curve misses some of the fluctuations captured by the red curve X3 , 4 even though the red curve clearly deviates from the other two processes . Gadgil et al . showed rigorously that the time evolution of the second moments of a discrete population evolving as a first-order reaction network system can be represented explicitly in terms of the eigenvalues and eigenvectors of the matrix that governs the evolution of the mean population dynamics [8] . We apply their general results to the specific example of a first-order transition network in two ways . First , we use the spectral decomposition of the stationary variance to establish our main stochastic shielding result . Second , their result on time varying systems allows us to obtain the decomposition of the power spectrum in terms of the eigenvalue spectral decomposition , shown in Eqs 64–66 . Consider an arbitrary first-order reaction network with graph Laplacian L and matrix B satisfying Eqs 32–36 ( see Methods ) . The fact that the stationary covariance matrix decomposes into a sum of contributions from each edge in the graph follows from a straightforward calculation that we describe in Lemma 1 and Theorem 1 . We defer the proof of Lemma 1 to §Methods , below . Definition 1 Let X denote the set of n × n real matrices C such that for all j = 1 , … , n , ∑ i = 1 n C i j = 0 . Let Y = {C ∈ X | C⊺ = C} . Lemma 1 Let L be an n × n real valued matrix with Lij ≥ 0 for i ≠ j , and Lii = −∑i , i ≠ j Lij ( so that ∑ i = 1 n L i j = 0 ) for j = 1 , … , n , and satisfying dim ( ker ( L ) ) = 1 , and with a null eigenvector Lv = 0 satisfying vi ≥ 0 for i = 1 , … , n . Then for any F ∈ Y , the equation L C + C L ⊺ = F ( 27 ) has a unique solution C ∈ Y . Theorem 1 For an arbitrary first-order reaction network with graph Laplacian L and matrix B satisfying Eqs 32–36 , there is a unique linear decomposition of the stationary covariance matrix C as a sum of contributions from each edge: C = ∑ k ∈ E C k w h e r e ( 28 ) C k = ∫ 0 ∞ ( e t L ) B k B k ⊺ ( e t L ) ⊺ d t ( 29 ) Proof 1 Proof of Theorem 1 . Consider a first-order reaction network defined by graph Laplacian L and matrix B , satisfying Eqs 32–36 . We want to solve the Lyapunov equation L C + C L ⊺ = - B B ⊺ ( 30 ) for matrix C . Note that L satisfies the conditions in Lemma 1 , and BB⊺ ∈ Y since BB⊺ is an n × n real symmetric matrix with columns that sum to zero . Then by Lemma 1 , Eq 30 has a unique solution C ∈ Y . By replacing F with BB⊺ in the proof of Lemma 1 , we see that the unique solution is C = ∫ 0 ∞ e t L B B ⊺ ( e t L ) ⊺ d t ( 31 ) since all eigenvalues of L have negative real part ( except for the Perron-Frobenius eigenvalue λ1 ≡ 0 ) , u 1 ⊺ B = 0 , and B⊺u1 = 0 . Since BB⊺ can be written as a sum of B k B k ⊺ , we can repeat the calculation above to get Eq 29 for each k separately . The integral in Eq 29 holds for all k since the kth stoichiometry vector ζk appearing in the kth column of B is orthogonal to the steady state eigenvector . Therefore , C decomposes into a sum over the Ck terms , and Eq 28 holds . The decomposition in Theorem 1 allows us to rank each edge in the network in terms of its contribution to the stationary variance of any given node , which we call its “importance” relative to that node . In the case of a single open or conducting node , we refer simply to the edge importance . Moreover , the decomposition allows us to quantify the accuracy of the stochastic shielding approximation with respect to the population process projected onto individual nodes . The decomposition given by Theorem 1 is exact regardless of timescale separation or node sparsity .
Markov chains provide a general framework for mathematically modeling and simulating stochastic processes in natural and artificial systems . However , Markov chains are computationally expensive as their simulations require random numbers at each time step for every transition ( edge ) . The stochastic shielding approximation relies on the fact that , when hidden states are present , the edges are not equally important , so that random fluctuations in some ( typically most ) edges can be neglected . Here , we provide a thorough study addressing how to identify the relevant and irrelevant edges when the stochastic fluctuations span slow and fast timescales . Our analysis shows that the stochastic shielding approach not only provides a practical increase in computational efficiency , but also facilitates a systematic understanding of the propagation of fluctuations in a general Markovian network , and hence , is applicable to many areas of mathematical biology and related disciplines . The stochastic shielding method is being used increasingly to incorporate fast , accurate simulation of stochastic ion channels into larger neuronal network models . A recent paper [35] comparing different methods for simulating ion channels , based on diffusion approximations , recommended using the stochastic shielding approximation in conjuction with a direct Langevin approach advanced by Orio and Soudry [36] . Two examples in which stochastic shielding makes large-scale simulations tractable include [37] and [38] . In the first paper , the use of stochastic shielding allowed for a significant reduction in computation time of multiple simulations of a conductance-based model with synaptic and ion channel noise that are necessary to reliably estimate the entropy and information rate of neuronal firing . In the second paper , stochastic shielding is applied to a heterogeneous neural circuit for the first time , allowing the authors to investigate the role of channel noise in the generation of breathing variability in the isolated central pattern generator of respiration . In both cases , these studies would have not been possible in practice without the stochastic shielding approximation . The analysis conducted here and in [10] is restricted to the case of a stationary Markov process , i . e . with time-invariant per capita transition rates . In many applications , for example under current-clamp ( rather than voltage-clamp ) in electrophysiology , the transition rates vary over time . In [9] , which introduced the stochastic shielding method , stochastic shielding was shown to produce accurate approximations through comparison of voltage traces and spike trains generated via both stochastic shielding and full Monte Carlo simulations . In the present paper , we have shown that in the presence of multiple timescales , for instance as seen in the dynamics of the nicotinic acetylcholine receptor ( nAChR ) under low agonist concentrations , one or more unobserved edges can become more important than the observable edges , in terms of making a greater contribution to the stationary variance of the occupancy of the open channel state ( and hence the variance of the ionic current through the population of channels ) . In such a case the stochastic shielding phenomenon is still present , but is significantly reduced , to the point that the approximation given by suppressing the noise on the hidden edges does not provide the best approximation . Indeed , as seen in Fig 10 , one may conclude that in this situation there is no suitable approximation of the type we consider , since the traces generated by reduced models with noise suppressed either on the observed or unobserved edges do not bear much similarity to the trace generated by the full model ( with identical noise forcing where the noise is included ) . On the one hand , the edge importance measure remains exact under all conditions , as long as the network is irreducible ( meaning here that α12 , α21 , α23 and α32 are all nonzero ) . On the other hand , the stationary variance does not capture the full shape of the trajectories . The decomposition of the fluctuations at one node as a sum of contributions from distinct edges extends to the correlation function and the power spectrum and the cross-spectrum , as well as to the variance . Motivated by the example of the nicotinic acetylcholine receptor , we systematically studied the effects of introducing separation of timescales into the simplest nontrivial model to which stochastic shielding applies: the 3-state chain with one observable state . We found that , in the case of two distinct timescales , accelerating or decelerating a subset of edges relative to a baseline case ( αij = 1 for all adjacent nodes ( i , j ) ) could in some cases enhance , and in other cases reduce the gap in edge importance between the observed and unobserved edges , but in no case could induce a reversal of the edge importance ( as observed in nAChR ) . Finally , by sampling an ensemble of different transition rates , we found that inversion of edge importance can be seen in a 3-state chain when the channel opening rate is large ( that is , α23 ≫ max ( α12 , α21 , α32 ) ) , and also the rate of return from the first hidden state to the middle hidden state is small ( that is , α12 ≪ α21 ) . These complementary conditions are captured by the exact expression for the relative edge importance ( Eq 22 ) . Together , these conditions lead to sparse occupation of the middle node , introducing a bottleneck , while also introducing timescale separation in such a way that equilibration between the observable node and its immediate neighbor occurs much faster than between the two unobservable nodes . Although our exact formula applies only to the 3-state chain model from which it was derived , we are optimistic that it may be extended to broader classes of Markov processes . The forms of such extensions are not a priori obvious , for several reasons . Consider the case of an ion channel with n states of which a single open conducting state ( On ) is connected to the closed , non-conducting states ( C1 , … , Cn−1 ) through a single bottleneck state ( Cn−1 ) ; the closed states may interconnect arbitrarily with rates αij , 1 ≤ i , j ≤ ( n − 1 ) . In this case the analog of the first factor in Eq 22 would be the conditional occupancy probability of the bottleneck node Cn−1 , given the channel is in any of the states C1 , … , Cn−1 . However , the analog of the second factor , the ratio of the Cn−1 → On transition to some combination of all the rates in the system , is far from clear . For ion channel models with multiple transitions into and out of a single open state ( see Fig 1 ) , the parallel to our exact 3-state chain analysis is scarcely obvious , and remains for future investigation . The stochastic shielding approximation and method provide an approach distinct from aggregation based on community structure [20] or similarity of spectral components [13 , 39] , and pruning of sparsely populated nodes [23 , 33] , although there are some relations between these methods . Both spectral coarse graining [13] and our edge importance measure [10] rely on spectral decomposition of the graph Laplacian . As Ullah et al . point out , finding eigenvalues and eigenvectors of the Laplacian for a large complicated graph can be challenging [23] . An advantage of the stochastic shielding method is that it can be applied in the vast majority of cases without calculating the edge importance explicitly . Exceptions can occur when there is significant timescale separation with fast relaxation of the observable node with its immediate neighbors and slow relaxation among unobservable states , with a hidden bottleneck state separating the observable from a well populated pool of unobservable nodes . Except in this particular case , the stochastic shielding method can be applied without necessarily having to calculate the edge importance in detail . The effect of fluctuations in rates along the hidden edges is filtered by the network , and their impact on fluctuations at the observable nodes is diminished .
We begin with a directed graph G = ( V , E ) with edge weights αij ≥ 0 representing a population of Ntot individuals moving randomly and independently among n states ( i , j ∈ V ) along m edges {i ( k ) →j ( k ) }1≤k≤m , with per capita transition rates {αk}1≤k≤m . We emphasize that edge k is the unique directed edge connecting source node i ( k ) to destination node j ( k ) . The n × 1 stoichiometry vector ζk corresponding to edge k is defined such that ζk ( i ) = −1 and ζk ( j ) = +1 , otherwise ζk ( l ) = 0; these vectors represent the effect of a transition along edge k . We use this notation to be consistent with the edge importance formula in the next subsection which is a sum of contributions to the variance of the observable state coming from each edge . Also , note that we will write the per capita transition rates either with double indexing denoting the source and destination nodes ( αij ) or with a single index denoting the reaction ( αk ) . We represent the population state at time t with an integer-valued vector N ( t ) = ( N1 ( t ) , … , Nn ( t ) ) ⊺ , where Ni ( t ) ≥ 0 and ∑ i = 1 n N i ( t ) = N tot for all t . In other words , N ( t ) is a discrete state continuous time Markov process . Such processes are ubiquitous in biology [1] . We denote by M a measurement vector indicating a direction in the state space along which there is an observable of interest . For instance , Mi ∈ {0 , 1} could denote the conducting state ( {closed , open} ) in a multi-state ion channel model . We denote the observed process by Y ( t ) = M⊺N ( t ) . The remainder of our set up follows standard nomenclature for representing a population process on a graph [8 , 40–42] . Let L be the Laplacian of graph G which is the n × n matrix defined by L = ( A − D ) ⊺ where A is the weighted adjacency matrix and D is the diagonal matrix of node out-degrees . Specifically , the entries in A are Aij = αij ( k ) = αk ≥ 0 and the diagonal entries in D are D i i = ∑ j = 1 n A i j for i ∈ {1 , … , n} . Note that L = Q⊺ where Q is the standard generator matrix of the Markov process . It follows that , for any vector x ∈ R n , L satisfies the following equation L x ≡ ∑ k = 1 m ζ k α k x i ( k ) . ( 32 ) The stoichiometry vector ζk is a difference of two standard unit vectors , ζk = ej ( k ) − ei ( k ) . Although we do not assume that the graph Laplacian L must be a symmetric matrix , we do assume that the stationary system satisfies detailed balance , and that L has only real eigenvalues . Moreover , we assume that L has an expansion into real-valued biorthogonal eigentriples ( wλ , λ , vλ ) such that L v λ = λ v λ ( 33 ) L ⊺ w λ = λ w λ ( 34 ) w λ ⊺ v λ ′ = δ λ λ ′ . ( 35 ) We further assume that G is connected and the process is irreducible . The Perron-Frobenius theory guarantees the existence of a unique null eigenvector with nonnegative components summing to unity , corresponding to the stationary distribution on the graph . We denote the stationary probability vector π → = ( π 1 , … , π n ) ⊺ and the stationary mean flux along edge k by Jk = Ntotαkπi ( k ) . Let B be the n × m matrix defined such that B = ( J 1 ζ 1 J 2 ζ 2 ⋯ J m ζ m ) . ( 36 ) In other words , the kth column of B is given by the square root of the stationary flux Jk multiplied by the stoichoimetry vector ζk . We can express B as a sum of matrices B = ∑ k = 1 m B k ( 37 ) where all the entries of Bk are zero except for the kth column . Moreover , we will exploit the fact that the product BB⊺ can be represented with a similar sum B B ⊺ = ∑ k = 1 m B k B k ⊺ . ( 38 ) This product appears on the right hand side of the Lyapunov equation ( see Eq 46 below ) and its decomposition into the above sum is a key factor in establishing the decomposition of the stationary variance into a sum over the edges . Computations were done either by hand , or using Matlab or Mathematica . In the Langevin approximation for a time homogeneous first-order transition network , the population fraction occupying states 1 , … , n is a vector X ∈ R n satisfying d X d t = L X + ∑ k ∈ E B k ξ k ( 39 ) where L , the graph Laplacian , captures the mean flux along each directed edge k ∈ E . The matrix Bk gives the effects of fluctuations ξk around the mean flux along the kth edge . The noise terms are independent , white and Gaussian , with 〈ξk ( t ) ξk′ ( t′ ) 〉 = δkk′ δ ( t − t′ ) , one for each directed edge . Given an observable of interest , represented by a vector M ∈ R n , the stochastic shielding approximation consists in finding a partition of the edge set , E = E 1 ∐ E 2 , into edges of primary importance ( E 1 ) and secondary importance ( E 2 ) such that | E 2 | ≫ | E 1 | and , at the same time lim t → ∞ E | | M ⊺ ( Y ( t ) - X ( t ) ) | | 2 ⪡ lim t → ∞ E | | M ⊺ X ( t ) | | 2 ( stationary variances ) , where Y is the approximate population vector satisfying d Y d t = L Y + ∑ k ∈ E 1 B k ξ k , Y ( 0 ) = X ( 0 ) . ( 40 ) The noise samples ξk for k ∈ E 1 are identical in the full and approximate models . Neglecting the noise forcing along the edges of secondary importance causes a pathwise discrepancy U ( t ) = Y ( t ) − X ( t ) that satisfies d U d t = L U - ∑ k ∈ E 2 B k ξ k , U ( 0 ) = 0 . ( 41 ) The stochastic shielding effect consists in suppression of the resulting fluctuations in the observable process M⊺U ( t ) due to the filtering effects of the network—hence “stochastic shielding” . The ( stationary ) mean squared pathwise approximation error can be written exactly as a sum of contributions Rk from each directed edge neglected in the approximation , lim t → ∞ E | | M ⊺ U ( t ) ) | | 2 = ∑ k ∈ E 2 R k . This error is small compared to lim t → ∞ E | | M ⊺ X ( t ) ) | | 2 = ∑ k ∈ E R k = ∑ k ∈ E 1 R k + ∑ k ∈ E 2 R k . We call Rk the importance of the kth directed edge ( defined in the next section ) . As we show below , the decomposition holds exactly not only for the Langevin process but for the discrete population process as well . The general formula for the edge importance measure is as follows . For an arbitrary stationary population process N ( t ) satisfying detailed balance on a graph with n nodes , m edges , and measurement vector M ( defining the observable states ) , R = ∑ k = 1 m R k is the stationary variance of the observable states where R k = J k ∑ i = 2 n ∑ j = 2 n ( - 1 λ i + λ j ) ( M ⊺ v i ) ( w i ⊺ ζ k ) ( ζ k ⊺ w j ) ( v j ⊺ M ) . ( 42 ) In this formula , λn ≤ λn−1 ≤ ⋯ ≤ λ2 < 0 are the nontrivial eigenvalues of the graph Laplacian L ( which always has λ1 ≡ 0 ) ; vi and wi are the right and left eigenvectors of L , respectively . Here and henceforth , Rk is normalized to the variance due to a single random walker by dividing out Ntot . The stationary variance R is related to the power spectral density ( PSD ) S ( ω ) of the observed process M⊺N . From the Wiener-Khinchin theorem , integrating the PSD gives the stationary variance: R = ∫ - ∞ ∞ S ( ω ) d ω . Moreover , since the stationary variance decomposes into a sum of contributions from each edge in the graph , the power spectral density decomposes as well . By introducing R k = ∫ - ∞ ∞ S k ( ω ) d ω ( 43 ) we define a power-spectral edge importance such that the integral of Sk ( ω ) , the power spectral density for the observed process with noise suppressed everywhere except edge k , gives the edge importance corresponding to edge k . To see this , note that the power spectral density of the observed process is S ( ω ) = ∑ k ∈ E S k ( ω ) where ( 44 ) S k ( ω ) = 1 2 π J k ∑ l = 2 n ∑ j = 2 n ( 1 λ l + i ω ) ( 1 λ j - i ω ) ( M ⊺ v l ) ( u l ⊺ ζ k ) ( ζ k ⊺ u j ) ( v j ⊺ M ) ( 45 ) provided ω > 0 . For more details , see §Numerical Methods: Calculation of power spectra , below . We can use this power spectral decomposition to explore how the spectral contributions differ between the typical cases ( where edge importance ranking agrees with the stochastic shielding method ) and in the edge importance reversal cases . The Perron-Frobenius null eigenvector , suitably normalized , gives the stationary probability vector π → = ( π 1 , … , π n ) ⊺ of Markov process N ( t ) . Snapshots of the process N ( t ) , taken under stationary conditions , are multinomial with parameters N tot , π → , so the covariance matrix C is known . In particular , each diagonal entry in C is the variance of state i , Cii = Ntotπi ( 1 − πi ) , and each off-diagonal entry in C is the covariance of states i and j , Cij = −Ntotπiπj for i ≠ j . The stationary covariance matrix C satisfies Lyapunov’s equation ( a special case of Sylvester’s equation ) [43] L C + C L ⊺ = - B B ⊺ . ( 46 ) The fact that C satisfies Eq 46 above is widely known for linear Gaussian processes such as multivariate Ornstein-Uhlenbeck processes [34] , but it also holds for discrete state population processes in which the transition rates are linear functions of the population at each node , i . e . first-order transition networks , such as those we consider here ( see [8 , 44] ) . Our system is an important special case of the general first-order reaction network presented in [8]; we only consider conversion type reactions ( denoted by kcon in [8] ) . For our system Pi represents v λ u λ ⊺ , summed over all identical λ if they occur with multiplicity ( we both assume semisimple eigenvalue spectra ) . The following parameters in [8] are zero for our system: C ( i , k , l ) , kcat , ks , and kd . This simplifies Equation 50 in [8] ( representing the variance of the lth reactant in the network ) and is equivalent to our edge importance measure ( Eq 42 ) . However , to our knowledge , we are the first to describe the unique decomposition of the stationary variance into a sum of contributions from each edge in the network , and [9 , 10] were the first to propose the stochastic shielding approximation and justify it based on this decomposition . The Lyapunov equation has also been used in the context of stochastic gene networks under the name of “linear noise approximation” [45 , 46]; in particular [45] ( pg . 1 , ¶5 ) further cites Eqs 3 . 46 and 6 . 115 in Risken [47] for additional details . See also [48] Supporting Information §4 . For the linear networks we consider here , the equation is exact . We restate the lemma for the reader’s convenience . Recall from Definition 1 that Y is the space of n × n symmetric matrices with columns ( and rows ) summing to zero . Lemma 1 ( restated ) Let L be an n × n real valued matrix with Lij ≥ 0 for i ≠ j , and Lii = −∑i , i ≠ j Lij ( so that ∑ i = 1 n L i j = 0 ) for j = 1 , … , n , and satisfying dim ( ker ( L ) ) = 1 , and with a null eigenvector Lv = 0 satisfying vi ≥ 0 for i = 1 , … , n . Then for any F ∈ Y , the equation L C + C L ⊺ = F has a unique solution C ∈ Y . Proof 2 Proof of Lemma 1 . Given L ∈ X , define the linear operator A by A: C → LC + CL⊺ . First , we show that A: Y → Y . If C ∈ Y then for all j = 1 , … , n , ∑i=1n ( LC+CL⊺ ) ij=∑i , k=1n ( LikCkj+CikLjk ) =∑k=1nCkj∑i=1nLik+∑k=1nLjk∑i=1nCik=0 , ( 47 ) because each sum over i is zero , by assumption . Moreover , ( LC + CL⊺ ) ⊺ = LC + CL⊺ . Therefore LC + CL⊺ ∈ Y whenever C ∈ Y , so A maps Y into itself . By the Fredholm alternative ( cf . [49] , Theorem 2 . 27 ) , A ( C ) = F has a unique inverse for F ∈ Y provided F is in the range of A and the homogeneous equation A ( C ) = 0 has only the trivial solution C = 0 . Let C0 ∈ Y be a solution of the homogeneous equation , LC0 + C0L⊺ = 0 . Because C0 ∈ Y is symmetric and the nullspace of L is one dimensional , C0 must have the form C0 = ( c1v|⋯|cnv ) for constants c1 , … , cn . However , the columns of C0 must sum to zero , and ∑ i = 1 n v i > 0 , therefore c1 = … = cn = 0 , hence C0 = 0 . To see that F is in the range of A , we construct an explicit solution as follows: C = ∫ 0 ∞ e t L F ( e t L ) ⊺ d t , ( 48 ) and we show that this integral is well defined whenever F ∈ Y . To see this , first note that if all eigenvalues of L have negative real part , then L C + C L ⊺ = ∫ 0 ∞ S d t ( 49 ) where S = L e t L F e t L ⊺ + e t L F e t L ⊺ L ⊺ ( 50 ) = d d t ( e t L F e t L ⊺ ) ( 51 ) and the solution in Eq 48 follows from the fundamental theorem of calculus . It remains to show that the integral in Eq 48 is well defined whenever F ∈ Y . Assuming detailed balance , a unique null space , and that L is diagonalizable , we have that all eigenvalues of L are negative ( and real ) except λ1 ≡ 0 , and we can write L = ∑ λ v λ u λ ⊺ ( 52 ) ⇒ e t L = v 0 u 0 ⊺ + ∑ λ < 0 e t λ v λ u λ ⊺ . ( 53 ) Then C = ∫ 0 ∞ e t L F ( e t L ) ⊺ d t ( 54 ) = ∫ 0 ∞ { ( v 1 u 1 ⊺ ) F ( v 1 u 1 ⊺ ) ⊺ + ( v 1 u 1 ⊺ ) F ( ∑ λ < 0 e t λ v λ u λ ⊺ ) ⊺ ( 55 ) + ( ∑ λ < 0 e t λ v λ u λ ⊺ ) F ( v 1 u 1 ⊺ ) ⊺ + ∑ λ < 0 , λ ′ < 0 e t ( λ + λ ′ ) v λ u λ ⊺ F u λ ′ v λ ′ ⊺ } d t ( 56 ) = ∫ 0 ∞ { v 1 ( u 1 ⊺ F _ ) ( u 1 v 1 ⊺ ) + v 1 ( u 1 ⊺ F _ ) ∑ λ < 0 e t λ u λ v λ ⊺ ( 57 ) + ∑ λ < 0 e t λ v λ u λ ⊺ ( F u 1 _ ) v 1 ⊺ + ∑ λ < 0 , λ ′ < 0 e t ( λ + λ ′ ) v λ u λ ⊺ F u λ ′ v λ ′ ⊺ } d t ( 58 ) = ∑ λ < 0 , λ ′ < 0 v λ u λ ⊺ F u λ ′ v λ ′ ⊺ ( ∫ 0 ∞ e t ( λ + λ ′ ) d t ) ( 59 ) = ∑ λ < 0 , λ ′ < 0 - 1 λ + λ ′ v λ u λ ⊺ F u λ ′ v λ ′ ⊺ . ( 60 ) The underlined expressions in parentheses are all zero because the columns ( and rows since F is a symmetric matrix ) of F sum to zero by assumption; u 1 ⊺ ≡ ( 1 , … , 1 ) is orthogonal to every column of F and u1 is orthogonal to every row of F and so u 1 ⊺ F = 0 and Fu1 = 0 . Thus , the integral in Eq 54 is finite and Eq 60 gives an explicit expression for it . | Discrete state , continuous time Markov processes occur throughout cell biology , neuroscience , and ecology , representing the random dynamics of processes transitioning among multiple locations or states . Complexity reduction for such models aims to capture the essential dynamics and stochastic properties via a simpler representation , with minimal loss of accuracy . Classical approaches , such as aggregation of nodes and elimination of fast variables , lead to reduced models that are no longer Markovian . Stochastic shielding provides an alternative approach by simplifying the description of the noise driving the process , while preserving the Markov property , by removing from the model those fluctuations that are not directly observable . We previously applied the stochastic shielding approximation to several Markov processes arising in neuroscience and processes on random graphs . Here we explore the range of validity of stochastic shielding for processes with nonuniform stationary probabilities and multiple timescales , including ion channels with “bursty” dynamics . We show that stochastic shielding is robust to the introduction of timescale separation , for a class of simple networks , but it can break down for more complex systems with three distinct timescales . We also show that our related edge importance measure remains valid for arbitrary networks regardless of multiple timescales . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"electrophysiology",
"random",
"variables",
"neuroscience",
"covariance",
"simulation",
"and",
"modeling",
"ion",
"channels",
"mathematics",
"statistics",
"(mathematics)",
"algebra",
"research",
"and",
"analysis",
"methods",
"nicot... | 2018 | Stochastic shielding and edge importance for Markov chains with timescale separation |
3% of the population develops saccular intracranial aneurysms ( sIAs ) , a complex trait , with a sporadic and a familial form . Subarachnoid hemorrhage from sIA ( sIA-SAH ) is a devastating form of stroke . Certain rare genetic variants are enriched in the Finns , a population isolate with a small founder population and bottleneck events . As the sIA-SAH incidence in Finland is >2× increased , such variants may associate with sIA in the Finnish population . We tested 9 . 4 million variants for association in 760 Finnish sIA patients ( enriched for familial sIA ) , and in 2 , 513 matched controls with case-control status and with the number of sIAs . The most promising loci ( p<5E-6 ) were replicated in 858 Finnish sIA patients and 4 , 048 controls . The frequencies and effect sizes of the replicated variants were compared to a continental European population using 717 Dutch cases and 3 , 004 controls . We discovered four new high-risk loci with low frequency lead variants . Three were associated with the case-control status: 2q23 . 3 ( MAF 2 . 1% , OR 1 . 89 , p 1 . 42×10-9 ) ; 5q31 . 3 ( MAF 2 . 7% , OR 1 . 66 , p 3 . 17×10-8 ) ; 6q24 . 2 ( MAF 2 . 6% , OR 1 . 87 , p 1 . 87×10-11 ) and one with the number of sIAs: 7p22 . 1 ( MAF 3 . 3% , RR 1 . 59 , p 6 . 08×-9 ) . Two of the associations ( 5q31 . 3 , 6q24 . 2 ) replicated in the Dutch sample . The 7p22 . 1 locus was strongly differentiated; the lead variant was more frequent in Finland ( 4 . 6% ) than in the Netherlands ( 0 . 3% ) . Additionally , we replicated a previously inconclusive locus on 2q33 . 1 in all samples tested ( OR 1 . 27 , p 1 . 87×10-12 ) . The five loci explain 2 . 1% of the sIA heritability in Finland , and may relate to , but not explain , the increased incidence of sIA-SAH in Finland . This study illustrates the utility of population isolates , familial enrichment , dense genotype imputation and alternate phenotyping in search for variants associated with complex diseases .
About 3% of the population develops saccular intracranial aneurysms ( sIAs ) during life [1] , [2] . Some 95% of subarachnoid hemorrhages are caused by ruptured sIA ( sIA-SAH ) , a devastating form of stroke affecting individuals mainly in the sixth decade of life [3] . The annual incidence of SAH is 4–9 per 100 000 worldwide [4] but over twice as high in Finland and in Japan [5] . The sIA disease is a complex trait , the risk of which is affected by age , sex , smoking , hypertension , excess drinking [6] , and in over 10% of the cases family history of sIA disease [7]–[9] . To date , genome wide association ( GWA ) studies have identified six definite and one probable loci with common variants associated to sIA: 4q31 . 23 ( OR 1 . 22 ) [10] , [11]; 8q11 . 23–q12 . 1 ( OR 1 . 28 ) ; 9p21 . 3 ( OR 1 . 31 ) ; 10q24 . 32 ( OR 1 . 29 ) ; 12q22 ( OR 1 . 16 ) [10]; 13q13 . 1 ( OR 1 . 20 ) ; 18q11 . 2 ( OR 1 . 22 ) [12] ( Table S5 ) . These seven loci were estimated to explain 6 . 1% , 4 . 4% and 4 . 1% of the four-fold sibling recurrence risk in Finland , Europe and Japan respectively [10] . In these previous GWA studies , results on 2q33 . 1 locus were inconsistent: the locus was significant in the first GWAS [13] , not significant in the enlarged follow-up GWAS [12] , and in the third GWAS the risk allele was reversed in the Japanese replication sample [10] . The population of Finland is one of the most thoroughly characterized genetic isolates . Due to the small size of the founder population , subsequent bottleneck effects and genetic drift , the Finnish population is enriched for rare and low frequency variants that are almost absent in other European populations and some variants rare elsewhere are increased in frequency [14] . This is best illustrated by the increased prevalence of 36 rare Mendelian , mostly recessive , disorders in Finland ( www . findis . org ) ; the so called Finnish disease heritage ( FDH ) [15] . We hypothesized that some of the enriched rare or low frequency variants could contribute to the increased sIA-SAH susceptibility in Finland . In this GWA study we combined the power of 1000 Genomes imputation , the special benefits of a population isolate and enrichment of familial cases in the discovery cohort . Familial sIA patients more often carry multiple sIAs as compared to sporadic sIA patients , which may confer additional genetic burden to the sIA formation [8] , [16] , [17] . Therefore , in addition to the case vs . control analysis , we also analyzed the number of sIAs per individual as an intermediate phenotype . We conducted an association analysis in a discovery sample of 760 Finnish sIA cases and 2 , 513 matched controls followed by replication in an additional sample of 858 Finnish sIA cases and 4048 controls . The successfully replicated loci in Finland were further studied in a Dutch cohort of 717 sIA cases and 3004 controls to assess the extent to which the allele frequencies and risk effect sizes match between the isolate of Finland and a continental European population ( Figure 1 ) . In addition , we hypothesized that a previously inconclusive locus on 2q33 . 1 [10] , [13] , [18] is a true sIA risk locus at least in Finland and aimed to replicate the best discovery associations in the locus in this study in the Finnish and in the Dutch samples . We successfully identified associations with low frequency variants in three novel loci in the case vs . control analysis and one in the aneurysm count analysis . Two of the case vs . control loci replicated also in the Dutch cohort with similar allele frequencies and comparable risk effect sizes . The variant in the aneurysm count locus demonstrated a strong bottleneck effect by being 15 times more frequent in the Finnish than in the Dutch controls . We also successfully replicated the previously inconclusive 2q33 . 1 locus .
To increase the potential genetic load in the study sample , our discovery sample consisted of 760 cases from the isolated , high-risk Finnish population , purposefully enriched for familial sIA ( 40% ) patients and 2513 genetically matched Finnish controls . The imputation of the 304 , 399 previously genotyped variants [12] against the 1000 Genomes Project reference panel ( v3 , March 2012 release ) increased the number of common and low frequency variants available for the association analysis to 9 , 359 , 231 . Quantile-quantile ( QQ ) plots of association p-values did not indicate substantial inflation ( λ = 1 . 04 ) ( Figure S1 ) . The discovery association analysis revealed one locus at 12p11 . 1 driven by rs653464 at conventional genome-wide significance ( p<5×10−8 ) and 14 other loci at p<5×10−6 ( Table S1; Manhattan plot in Figure S3 ) . We chose 17 SNPs representing the 15 promising loci ( p<5×10−6 ) above for replication in an independent sample of 858 Finnish sIA cases and 4 , 048 controls ( Table 1 ) . Four SNPs and one deletion were associated at p<0 . 05 with the sIA disease ( Table S1 ) , two of them in the previously reported sIA loci 9p21 . 3 ( rs1333042; OR 1 . 3 , p = 6 . 3×10−7 ) and 13q13 . 1 ( rs113124623; OR 0 . 88 , p = 0 . 01 ) . The genome-wide significant 12p11 . 1 locus in the discovery sample did not replicate ( p = 0 . 29 ) . In the meta-analysis of the two Finnish samples , four SNPs reached the commonly used level of genome-wide significance at p<5×10−8 ( Table 2 ) . Three were novel: 2q23 . 3 ( rs74972714; OR 2 . 1 , 95% CI 1 . 68–2 . 63 , p = 7 . 4×10−11 , control allele frequency or CAF 2 . 35% ) , 5q31 . 3 ( rs113816216; OR 1 . 92 , CI 1 . 53–2 . 40 , p = 1 . 74×10−8 , CAF 2 . 09% ) and 6q24 . 2 ( rs75018213; OR 1 . 97 , CI 1 . 6–2 . 43 , p = 2 . 25×10−10 , CAF 2 . 53% ) . One was previously reported at 9p21 . 3 ( rs1333042; OR 1 . 31 , CI 1 . 21–1 . 42 , p = 1 . 8×10−11 , CAF 42 . 3% ) ( Table 2 ) . We assessed the robustness of the associations controlling also for age and the effect sizes and p-values were almost identical ( data not shown ) . To assess how the allele frequencies and effect sizes of variants identified in the Finnish population compare to other European populations , we studied those variants in a Dutch sample consisting of 717 sIA cases and 3 , 004 controls ( Table 1 ) . All three variants tagging the novel loci at 2q23 . 3 , 5q31 . 3 and 6q24 . 2 had a similar low minor allele frequency ( 1 . 6–3 . 9% ) in Finland and the Netherlands ( Table 2 ) . Two of them had similar effect sizes and were also replicated: 5q31 . 3 ( rs113816216; OR 1 . 3 , CI 0 . 98–1 . 75 , p = 0 . 045 , CAF 3 . 87% ) and 6q24 . 2 ( rs75018213; OR 1 . 5 , CI 0 . 98–2 . 3 p = 0 . 034 , CAF 2 . 3% ) . The previously reported 9p21 . 3 locus also replicated in the Dutch sample ( rs1333042; OR 1 . 32 , CI 1 . 17–1 . 49 , p = 3 . 42×10−6 , CAF 47 . 86% ) . In the meta-analysis of the Finnish and Dutch samples , all three novel loci 2q23 . 3 ( rs74972714; OR 1 . 89 , p = 1 . 42×10−9 ) , 5q31 . 3 ( rs113816216; OR 1 . 66 , p = 3 . 17×10−8 ) and 6q24 . 2 ( rs75018213; 1 . 87 , p = 7 . 1×10−11 ) were significantly associated to the sIA disease at genome-wide significance ( Table 2; see Table S7 for imputation accuracy statistics ) . Some heterogeneity in effect sizes exists between samples ( Table S9 ) . As the standard genome-wide significance 5×10−8 is estimated to correct for independent tests of common variants ( MAF> = 5% ) and we tested also a set of low-frequency variants , the common significance level may be too liberal . Based on Europeans of the 1000 Genomes project we estimated the significance level to be 3 . 82×10−8 ( See Materials and Methods ) . All of the reported variants are below this level . Some 20–30% of the sIA patients carry multiple sIAs , a phenomenon more commonly seen in familial sIA disease [8] , [16] , [17] . We hypothesized that an increased number of sIAs ( ≥2 ) in a given patient would reflect a higher underlying genetic load , motivating us to use aneurysm count as an intermediate phenotype to increase statistical power . The number of sIAs was used as a count data using the negative binomial regression analysis in the discovery sample of 760 Finnish sIA cases ( 1–8 sIAs per patient ) and 2 , 513 controls . The QQ plot ( Figure S2 ) and the genomic inflation factor ( 1 . 05 ) did not indicate substantial population stratification . Nine loci had variants at p<5E-6 ( Table S2; Manhattan plot in Figure S4 ) . The most significant variant of each locus was selected for replication in the new Finnish sample of 858 sIA cases ( 1–6 sIAs per patient ) and 4 , 048 controls . Two loci were replicated at p<0 . 05: 7p22 . 1 ( rs150927513; RR 1 . 39 , p = 8 . 36×10−4 , CAF 5 . 24% ) and 16p13 . 3 ( rs144159053; rate ratio ( RR ) 1 . 66 , p = 4 . 4×10−3 , CAF 1 . 27% ) ( Table S2 ) . rs10802056 on 1p12 had a significant association p-value but the effect direction was different and thus was not considered as replicated . We assessed the robustness of the associations controlling also for age and the effect sizes and p-values were almost identical ( data not shown ) . In the meta-analysis of the Finnish samples , 7p22 . 1 was genome-wide significant ( rs150927513; RR 1 . 6 , CI 1 . 37–1 . 88 , p = 4 . 92×10−9 , CAF 4 . 61% ) ;Table 3; See genotype to aneurysm count distribution in Table S3 ) . The rate ratio ( RR ) estimate is the relative rate of aneurysm formation ( i . e . change in expected number of aneurysms ) per allele as compared to minor allele homozygotes . To compare the allele frequency and effect size of rs150927513 identified in the Finnish population to those of continental European populations , we studied the variant also in the Dutch , but the imputation quality ( Impute info 0 . 38 ) and estimated allele frequency ( 0 . 29% ) were too low to obtain reliable estimates ( RR 0 . 97; 95% CI 0 . 17–4 . 03 , p = 0 . 97 ) . We additionally checked the minor allele frequency of rs150927513 in 498 whole-genome sequenced Dutch individuals of GENOMEoftheNETHERLANDS-project ( http://www . nlgenome . nl/ ) . Only two individuals were heterozygous and the rest were major allele homozygotes ( MAF 0 . 2% ) , which is in agreement with our imputation results of the Dutch sample . Previously published results on the 2q33 . 1 locus are inconsistent , being significant in the first GWAS [13] , not significant in the enlarged follow-up GWAS [12] , and uncertain in the third GWAS [10] . We aimed to study if the 2q33 . 1 would replicate in Finland , even though no variant in this region reached p<5E-6 in the discovery sample . We chose two of the most significant SNPs ( in this study ) at 2q33 . 1 for replication in the new Finnish replication sample , which was not used in the previous studies ( rs12472355; OR 1 . 21 , p = 2 . 23×10−4 , CAF 44 . 3% , and rs919433; OR 1 . 18 , p = 1 . 01×10−3 , CAF 44 . 6% ) . They are in LD with the three previously investigated SNPs ( rs787994 , rs1429412 , rs700651; LD r2 0 . 75–0 . 96 ) . The variants rs12472355 ( OR 1 . 23 , CI 1 . 13–1 . 33 , p = 4 . 84×10−7 ) and rs919433 ( OR 1 . 21 , CI 1 . 12–1 . 31 , p = 2 . 15×10−6 ) did not reach genome-wide significance in the combined Finnish samples ( Table 2 ) . They were highly significant in the Dutch sample ( rs12472355; OR , 1 . 39 , CI 1 . 23–1 . 57 , p = 1 . 05×10−7 and rs919433; OR 1 . 43 , CI 1 . 26–1 . 61 , p 9 . 77×10−9 ) , and in the meta-analysis of all three samples they reached genome-wide significance ( Table 2 ) . The allele frequencies were notably higher in the Finnish samples ( 44% and 43 . 7% ) than in the Dutch samples ( 33 . 2% and 31% ) . We estimated the heritability explained by the reported variants . The four novel loci on 2q23 . 3 , 5q31 . 3 , 6q24 . 2 and 7p22 . 1 were estimated to explain 1 . 7% of the heritability in the combined Finnish samples . Adding the previously inconclusive 2q33 . 1 locus increases the heritability explained to 2 . 1% . For validating the imputation accuracy , we genotyped 87 individuals of the discovery sample using Sequenom genotyping . The concordance rates range from 96–99% except rs74972714 was slightly lower at 94% ( Table S8 ) . We did additional validation by Sanger sequencing 10 individuals per variant who were predicted to carry minor alleles . The imputation was near perfect in all other SNPs except rs75018213 had discrepancies between major allele homozygote and heterozygotes ( Table S11 ) . We further estimated by simulation , how likely it would be to get the observed OR for rs75018213 in the discovery sample just by change , given the imputation accuracy ( See Text S1 for details ) . The probability of chance finding was very low ( p: 0 . 0001 ) even if assuming that the minor allele would be over-imputed by 20% in cases ( p: 0 . 004 ) . Some individuals were genotyped by both Sanger sequencing and Sequenom and the concordance between the two methods was perfect ( Table S11 ) . Finally , we estimated , in silico , the imputation efficiency of reported SNPs in Dutch population . 96 individuals of the Genome of the Netherlands project had both high coverage whole-genome sequencing ( 40× ) data as well as GWA chip genotyping data available . We imputed the genotypes of reported SNPs using the same imputation methods , 1000 Genomes reference panel and set of SNPs in GWA chips as was done in the discovery and Dutch comparison analyses . The genotype concordance rates were excellent ( Table S13 ) . It is noteworthy that the imputation quality measure reported by the Impute2 program was higher in all of the SNPs in our Dutch replication cohort ( Table S7 ) than in the in silico validation experiment . This indicates excellent imputation quality in the Dutch replication . We attempted to identify putative causative variants from whole exome sequencing data of 583 Finnish individuals . We focused on variants within 1 MB of the lead SNPs with high impact on protein product ( i . e . variants affecting splice site , losing or gaining stop/start codon , altering reading frame ) or non-synonymous coding SNPs . We additionally filtered variants if they were not in LD with the lead SNPs ( r2<0 . 4 , Europeans of 1000 Genomes if available ) . 254 variants were identified , most of which were rare . However 15 variants were enriched to low-frequency range ( MAF>1% ) ( Table S12 ) . The impact of these variants needs to be evaluated in follow-up studies . The UCSC Genome Browser and HaploReg version 2 [19] were used to search for ENCODE regulatory elements at the five genome-wide significant variants . rs74972714 at 2q23 . 3 and rs150927513 at 7p22 . 1 reside within a DNAse hypersensitivity peak . The rs75018213 at 6q24 . 2 resides on an ENCODE GATA2 transcription factor binding site peak ( Table S4 ) . Using genome-wide Chip-SEQ analysis , Ernst et al . constructed a predicted cell-type specific regulatory region map of nine chromatin marks in nine cell lines [20] . rs113816216 at 5q31 . 3 resides on a predicted erythroleukemia cell specific ( K562 ) strong enhancer and rs75018213 at 6q24 . 2 on a predicted lymphoblastoid cell ( GM12878 ) weak enhancer ( Table S4 ) . We searched for putative transcription factor binding sites affected by the four variants , based on position weight matrices from Transfac , Jaspar and ENCODE ( top 3 enriched motifs for each transcription factor , identified in transcription factor Chip-SEQ peaks [19] ) . rs74972714 at 2q23 . 3 affects putative binding sites for EBF1 ( ENCODE ) , HDAC2 ( ENCODE ) , RXRA:PPARG complex ( Transfac ) , ZNF423 ( Jaspar ) and ZIC3 ( Jaspar ) . rs113816216 at 5q31 . 3 affects the putative binding sites for RFX1 ( Transfac ) , SREBP1 ( ENCODE ) , STAT3 ( Transfac ) and IKZF3 ( Transfac ) . rs150927513 at 7p22 . 1 affects putative binding sites of T ( brachyury ) ( Transfac ) , CEBPB ( Transfac ) and P300 ( ENCODE ) . rs75018213 at 6q24 . 2 is not directly on any putative transcription factor binding site . ( Table S4 ) . At the 2q33 . 1 locus neither of the studied variants ( rs919433 , rs12472355 ) are on ENCODE DNAse hypersensitivity or transcription factor binding site peaks . However , rs919433 is on a predicted lymphoblastoid ( GM12878 ) cell enhancer whereas rs12472355 is not directly on any regulatory region . rs919433 disrupts a putative transcription factor binding sites for RUNX2 ( OSF2 , Transfac ) and the MYC:MAX complex ( Transfac ) . To study the potential effects of the variants in the five significant loci on the transcripts of nearby genes , we correlated the variants to expression levels of exons of nearby genes ( expression quantitative trait locus ( eQTL ) analysis ) obtained using RNA-sequencing in lymphoblasts of genotyped European individuals from the 1000 Genomes Project ( Finnish , British , Toscani and CEPH populations , n = 373; www . geuvadis . org , [21] ) . Each variant was correlated to transcripts residing within 1 MB . There were 55 genes in 586 exons available for analysis ( see Materials and Methods ) and in total 748 tests were performed corresponding to Bonferroni corrected significance threshold of 8 . 7×10−5 . Strongest association for each variant are reported below and all eQTL results in Table S6 . The most significant eQTL associations were observed at the 2q33 . 1 locus: rs12472355 associated significantly to the closest gene ANKRD44 ( per allele fold change ( FC ) 0 . 94 , p = 1 . 83×10−5 ) and also to HSPD1 ( FC 0 . 94 , p = 1 . 6×10−4 ) , whereas rs919433 was associated to the same genes but in different order of significance; HSPD1 ( FC 0 . 94 , p = 3 . 8×10−5 ) and ANKRD44 ( FC 0 . 95 , p = 1 . 4×10−4 ) . Among the novel low-frequency variants , only rs150927513 at 7p22 . 1 was significantly associated to TNRC18 ( FC 1 . 23 , p = 5 . 1×10−5 ) . Nominal associations were observed for two other novel low frequency variants: rs113816216 at 5q31 . 3 to VDAC1 ( FC 2 . 12 , p 4 . 6E-4 ) ; rs74972714 at 2q23 . 3 to EPC2 ( FC 0 . 75 , p = 3 . 9×10−2 ) . rs75018213 at 6q24 . 2 did not have any association even at nominal p<0 . 05 ( Table S6 ) . We additionally investigated the eQTL landscape of identified loci by pairwise comparison of p-values from eQTL ( MAF>0 . 05 p<0 . 001 ) and sIA analyses ( Figure S5 ) and by plotting eQTL associations ( p<0 . 001 ) in the implicated loci ( Supplementary Figure S6 A–E ) . Only few loci show strong ( p<1E-5 ) association in eQTL and at least nominal ( p<0 . 05 ) association to sIA ( Table S10 ) . There does not seem to be stronger eQTL associations in LD with the lead SNPs . In the 2q33 . 1 , where the lead SNPs were significantly associated to transcript levels , there seems to be a lot of regulatory potential in the same locus , even though not in direct LD with the lead variants ( Figure S6 E ) .
The lead SNPs in the four novel loci all have a low frequency ( <5% ) in the general population and could not have been identified without imputing the genotype data against the 1000 Genomes reference . One of the variants , rs150927513 at 7p22 . 1 that was associated with the number of sIAs , indicates a strong bottleneck effect , for it was 15 times more frequent in the controls of combined Finnish samples ( 4 . 6% ) than in the Dutch sample ( 0 . 3% ) , and it is virtually non-existent in other populations ( 1000 Genomes ) . The three other loci had similar frequencies in Finland and other European populations ( 1000 Genomes ) . These four novel loci explain 1 . 7% of the heritability in the Finnish samples . The four sIA loci had higher effect sizes ( point estimates ranging from 1 . 59 to 1 . 88 ) than the lead SNPs identified by previous GWA studies . We cannot yet conclude whether relatively high ORs of low frequency risk alleles are a typical feature of sIA disease . Similar , and higher , odds ratios for low frequency and rare variants have been reported in isolates for other traits [22] , [23] . It is likely that this first wave of low frequency and rare susceptibility variants represent “low hanging” fruits that do not allow general conclusions about the susceptibility landscape of sIA or other complex traits . The variant rs74972714 at 2q23 . 3 has a frequency of about 2% in European populations , including Finns . It was significantly associated to sIA in the Finnish samples but did not show a trend for being associated in the Dutch sample despite having a similar allele frequency . Further studies are required to find out whether this variant tags a risk allele specific to Finnish sIA patients . The variant is located 40 kb downstream of LYPD6 and 55 kb upstream of MMADHC ( Figure 2 A ) . LYPD6 has recently been characterized as a member of the Ly-6 protein superfamily [24] . LYPD6 is ubiquitously expressed with highest levels in heart and brain . GPI-anchored Ly-6 proteins such as PLAUR function , e . g . , in cellular adhesion [24] . LYPD6 overexpression can inhibit transcriptional activity of the AP1 transcription factor complex [24] , a key inflammation mediator activated , e . g . , in endothelial cells in atherogenic disturbed blood flow conditions , leading in turn to upregulation of pro-inflammatory molecules [25] . Similar transcriptional changes have been found in the ruptured human sIA wall [26] . MMADHC is an intracellular vitamin B12 trafficking gene . Mutations in this gene can cause methylmalonic aciduria or homocystinuria , or both [27] . The variant rs113816216 at 5q31 . 3 has a frequency of 1–3% in Finland and most other European populations , except in Spain ( 7% ) . It was significantly associated to the sIA disease in the Finnish samples and was also significant in the Dutch sample but had a somewhat lower OR there ( Table 2 ) . The meta-analysis of all combined samples exceeded the genome wide significance threshold . The variant is located in the intron of FSTL4 ( Figure 2 B ) , a poorly characterized gene . FSTL1 , a paralog of FSTL4 , codes a protein inducing innate immunity as TLR4 agonist [28] . Increased tissue levels of FSTL1 were associated to the severity of heart failure [29] and to the coronary artery aneurysm formation in Kawasaki disease [30] . Variants in FSTL4 were modestly associated to human ischemic stroke [31] , and a variant 70 kb from FSTL4 nominally to hypertension [32] . The variant rs75018213 at 6q24 . 2 has similar frequencies ( 2% ) in European populations , including Finns . It was significantly associated to the sIA disease in the Finnish samples and was also significant in the Dutch sample but had a somewhat lower OR there ( Table 2 ) . It is located in the intron of EPM2A . The LD spans over 300 kb downstream covering FBXO30 , LOC100507557 , SHPRH and GRM1 ( Figure 2 C ) . In the ENCODE data , rs75018213 is located in a GATA2 transcription factor binding site RNA-seq peak . Homozygous deletions in the EPM2A gene result in progressive myoclonus epilepsy ( PME ) with Lafora bodies ( OMIM 254780 ) [33] . No vascular anomalies have been reported in EPM2 deletion patients with a PME phenotype or their heterozygote parents . EPM2A encodes a phosphatase , which dephosphorylates glycogen , but it is likely that EPM2A has broader functions in regulation of glycogen biosynthesis , endoplasmic reticulum stress , autophagy , and possibly also cell cycle [34] . The variant rs150927513 at 7p22 . 1 was significantly associated to sIA count per individual in the Finnish population ( Table 1 ) . Its frequency was 4 . 6% in the Finnish samples but only 0 . 3% , in the Dutch sample , in line with most European populations . This variant would therefore likely not have been identified if a sufficient number of Finnish individuals had not been included in the reference panel . The variant is located in the intron of RADIL ( Figure 2 D ) , a rap GTPase interactor , an essential effector of RAP1 in activation of integrins in cell-adhesive signalling by G protein-coupled receptors [35] . RADIL has also been shown to control , together with RAP1 , neutrophil adhesion and chemotaxis [36] . Neutrophils seem to have a role in the formation and rupture of intracranial and abdominal aortic aneurysm [26] , [37] , [38] . The strongest eQTL association was to an exon of TNRC18 ( FC 1 . 23 , p = 5 . 1×10−5 ) , a functionally uncharacterized gene . As we analysed the number of sIAs as a count variable from 0–8 , the inherent assumption was that the same variant would increase the risk of the first and the subsequent sIA formation . Thus , any variant associated to the number of sIAs will to some extent be associated in the case vs . control analysis . Indeed , in the analysis of combined Finnish cohorts rs150927513 was associated in the case-control analysis ( OR 1 . 54 , p = 6 . 5×10−7 ) and consistently also in the analysis of multiple vs . single sIA patients ( OR 1 . 65 , p = 8 . 4×10−4 ) . The association of this variant , should be interpreted as reflecting the tendency of sIA formation , rather than considering multiple sIAs as a completely different dichotomous end point . The 9p21 . 3 locus has been robustly associated to the sIA disease [12] as well as to cardiovascular , metabolic and cancer traits [39] , [40] , and it has been extensively studied by others [41] . The allele frequency and effect size in the current study , although with a different lead SNP ( r2 = 0 . 7 to previous lead SNP rs1333040 ) , are in strong agreement with the previous study [12] . This locus is not therefore discussed further here . Two common variants , rs12472355 and rs919433 at 2q33 . 1 were significantly associated to the sIA disease in the Finnish and Dutch samples ( Table 2 ) , rs919433 intronic and rs12472355 upstream 30 kb from ANKRD44 ( Figure 2 E ) . The allele frequencies were somewhat higher in the Finnish samples ( rs919433 , 44%; rs12472355 43 . 7% ) than in the Dutch samples ( 33 . 2%; 31% ) or in the Japanese population according to 1000 Genomes Project ( 28 . 1%; 27 . 5% ) . In this locus , the risk allele was reversed in the Japanese cohort of the previous sIA GWA study [10] . ANKRD44 is likely a subunit of protein phosphatase 6 [42] that functions , e . g . , in cell cycle control [43] and in inhibition of NF-κB activation [44] . NF-κB is a significant mediator in experimental sIA formation in rats , highly expressed in human sIA wall [45] , and it was associated to human sIA wall rupture in transcriptomic profiling [26] . In eQTL analysis rs12472355 was significantly associated to ANKRD44 ( FC 0 . 94 , p = 1 . 83×10−5 ) and rs919433 to HSPD1 ( FC 0 . 94 , p = 3 . 8×10−5 ) In conclusion , we identified four novel loci associated to sIA disease and confirmed one additional locus with previously inconclusive evidence , together explaining 2 . 1% of the sIA heritability in Finland . Our data illustrates the utility of high-risk population isolates , familial disease history , and dense genotype imputation in search for low-frequency variants associated to complex human diseases . The inclusion of Finnish individuals in the imputation reference panel and especially the highly differentiated variant in 7p22 . 1 would likely not have been identified The identification of the four novel low frequency variants would likely have required much larger sample sizes in more mixed populations . Further studies of the identified five loci are needed to explain their functional mechanisms in the pathogenesis of sIA disease .
For all of the Finnish and Dutch samples , the local ethics committees approved the study and all patients gave written informed consent . From both of the analyses ( the case vs . controls and the number of sIAs ) the best independent SNPs were taken to replication if p<5E-6 . Additional significant independent SNPs in a locus was tested by analyzing each SNP within 1 MB from the top SNP while adding the top SNP as a covariate . Additionally the most significant SNP in the current study in 2q33 . 1 region with uncertain evidence in previous sIA GWASs was taken to replication . Variant was considered replicated if it reached one-tailed significance of p<0 . 05 and was consistent in terms of risk allele . In all of the results , one-tailed p-values are given for the Finnish replication and in Dutch results . Genomic DNA was extracted from peripheral blood and genotyped by Illumina arrays: the Finnish discovery sample and the Dutch replication cases by CNV370k DUO chip; the HBCS and YFS controls by Illumina Human670K customBeadChip; and the H2000 controls by Illumina Infinium HDHuman610-Quad BeadChip . In the Finnish replication sample , DNA was genotyped using Sequenom MassARRAY system and iPLEX Gold assays ( Sequenom Inc . , San Diego , USA ) . The data was collected using the MassARRAY Compact System ( Sequenom ) and the genotypes were called using TyperAnalyzer software ( Sequenom ) . Genotyping quality was examined by a detailed QC procedure consisting of success rate checks , duplicates , water controls and Hardy-Weinberg Equilibrium ( HWE ) testing . SNPs were filtered if genotype missingness >0 . 05 or if HWE p<0 . 001 . For imputation of additional genotypes in the discovery sample , the Young Finns replication cohort and in the 2nd Dutch replication sample the genotypes were first pre-phased [55] using the Shape-IT [56] phasing software and the pre-phased haplotypes were subjected to imputation . The Impute version 2 . 2 . 2 software [57] with 1 , 000 Genomes Phase I integrated variant set release ( v3 ) reference panel ( 05 Mar 2012 release downloaded from http://mathgen . stats . ox . ac . uk/impute/data_download_1000G_phase1_integrated . html ) was used . Imputed genotypes were filtered if the Impute info measure was <0 . 5 or minor allele frequency <0 . 01 in the Finnish discovery sample . We analyzed whether the identified genome-wide significant SNPs might affect gene expression by using the European samples of the Geuvadis RNA-sequencing data set , with mRNA sequencing data from LCLs of 373 samples from the FIN , CEU , GBR and TSI populations of 1000 Genomes project ( for details , see [21] ) . We did eQTL analysis for each of the associating variants and all the genes within a 1 MB window that were expressed in >50% of the individuals ( Table eQTL ) . We used exon quantifications based on individual read counts per exon , after correction by the total number of mapped reads per sample and PEER normalization to remove technical variation . For each exon , we calculated linear regression between these expression values and genotype dosage of the associating variants in the 1000 Genomes data . Regional association plots were generated using LocusZoom with LD data from European populations of 1000 Genomes project ( Hg19/March 2012 ) [58] . The UCSC Genome Browser and HaploReg version 2 [19] were used to search for ENCODE regulatory element regions located at the five genome-wide significant variants . HaploReg database also annotates if SNP resides on a putative transcription-factor binding site ( TFBS ) according to Transfac or Jaspar TFBS profiles and also 10 most enriched TFBS profiles identified in ENCODE TF Chip-Seq peaks . We used all the Jaspar and Transfac annotations and three most enriched ENCODE based TFBS annotations for each TF . GWA was performed against two complementary phenotypes: the case vs . control status and the number of sIAs . | Genome-wide association studies ( GWAS ) have been extensively used to identify common genetic variants associated with complex diseases . As common genetic variants have explained only a small fraction of the heritability of most complex diseases , there is a growing interest in the role of how low frequency and rare variants contribute to the susceptibility . Low frequency variants are more often specific to populations of distinct ancestries . Saccular intracranial aneurysms ( sIA ) are balloon-like dilatations in the arteries on the surface of the brain . The rupture of sIA causes life-threatening intracranial bleeding . sIA is a complex disease , which is known to sometimes run in families . Here , we utilize the recent advancements in knowledge of genetic variation in different populations to examine the role of low-frequency variants in sIA disease in the isolated population of Finland where sIA related strokes are more common than in most other populations . By studying >8000 Finns we identify four low-frequency variants associated with the sIA disease . We also show that the association of two of the variants are seen in other European populations as well . Our findings demonstrate that multiple study designs are needed to uncover more comprehensively their genetic background , including population isolates . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"genome-wide",
"association",
"studies",
"stroke",
"genetics",
"biology",
"human",
"genetics",
"cardiovascular"
] | 2014 | High Risk Population Isolate Reveals Low Frequency Variants Predisposing to Intracranial Aneurysms |
Blood donor screening leads to large numbers of new diagnoses of Trypanosoma cruzi infection , with most donors in the asymptomatic chronic indeterminate form . Information on electrocardiogram ( ECG ) findings in infected blood donors is lacking and may help in counseling and recognizing those with more severe disease . To assess the frequency of ECG abnormalities in T . cruzi seropositive relative to seronegative blood donors , and to recognize ECG abnormalities associated with left ventricular dysfunction . The study retrospectively enrolled 499 seropositive blood donors in São Paulo and Montes Claros , Brazil , and 483 seronegative control donors matched by site , gender , age , and year of blood donation . All subjects underwent a health clinical evaluation , ECG , and echocardiogram ( Echo ) . ECG and Echo were reviewed blindly by centralized reading centers . Left ventricular ( LV ) dysfunction was defined as LV ejection fraction ( EF ) <0 . 50% . Right bundle branch block and left anterior fascicular block , isolated or in association , were more frequently found in seropositive cases ( p<0 . 0001 ) . Both QRS and QTc duration were associated with LVEF values ( correlation coefficients −0 . 159 , p<0 . 0003 , and −0 . 142 , p = 0 . 002 ) and showed a moderate accuracy in the detection of reduced LVEF ( area under the ROC curve: 0 . 778 and 0 . 790 , both p<0 . 0001 ) . Several ECG abnormalities were more commonly found in seropositive donors with depressed LVEF , including rhythm disorders ( frequent supraventricular ectopic beats , atrial fibrillation or flutter and pacemaker ) , intraventricular blocks ( right bundle branch block and left anterior fascicular block ) and ischemic abnormalities ( possible old myocardial infarction and major and minor ST abnormalities ) . ECG was sensitive ( 92% ) for recognition of seropositive donors with depressed LVEF and had a high negative predictive value ( 99% ) for ruling out LV dysfunction . ECG abnormalities are more frequent in seropositive than in seronegative blood donors . Several ECG abnormalities may help the recognition of seropositive cases with reduced LVEF who warrant careful follow-up and treatment .
Chagas disease ( ChD ) , caused by a flagellate protozoon , Trypanosoma cruzi ( T . cruzi ) , is a major health problem in Latin America , where more than 8 million persons are infected [1] , [2] . Chronic cardiopathy is the most important and severe manifestation of human Chagas disease , eventually affecting approximately 20% to 40% of those in the chronic phase of the disease [1] , [2] . Due to migratory movements , ChD is now a world-wide challenge , since hundreds of thousands of chronically infected persons are now living not only in T . cruzi endemic countries but also in developed countries , mainly in Europe and the United States and Canada [3] , [4] . Since one of the mechanisms of transmission of the disease is via blood transfusions , universal blood bank screening for ChD has been established in most endemic countries , as part of South American regional initiatives of elimination of transmission of the disease [5] , [6] . Non-endemic countries with large immigrant populations including the United States , Canada , Spain and Portugal , have also begun to institute interventions to prevent blood-borne T . cruzi transmission [7] . Blood donor antibody screening results in large numbers of new diagnoses of chronic T . cruzi infection , most of them in the asymptomatic , indeterminate form of infection [8] , [9] . Counseling these individuals should address the recognition of those with more severe disease that deserve to be rigorously evaluated by experienced cardiologists and treated more promptly . Electrocardiogram ( ECG ) , one of the most important tests in evaluation of ChD , is used to define the clinical stage of the disease with potential prognostic implications [10] . Most ECG studies of newly diagnosed ChD patients were performed decades ago , generally in patients identified by community or hospital based sampling; information on ECG findings in seropositive blood donors is lacking as is data relative to matched seronegative controls evaluated in parallel [11]–[17] . Additionally , most studies of ECG findings are not accompanied by systematic results that use core-lab reading of Echo and ECG results and codification by internationally accepted criteria , as the Minnesota Code for ECG findings [18] . As part of the National Heart , Lung and Blood Institute ( NHLBI ) Retrovirus Epidemiological Donor Study-II ( REDS-II ) , we developed a study to evaluate the prevalence of ECG abnormalities in seropositive blood donors and to recognize typical ECG abnormalities associated with left ventricular dysfunction , the most important prognostic marker in ChD .
This study was conducted from July 2008 to October 2010 , part of a retrospective cohort study in which T . cruzi seropositive blood donors identified by blood bank screening and well-matched seronegative donors . Using blood donation records from 1998–2002 , we enrolled 500 T . cruzi seropositive subjects ( 250 from the city of São Paulo and 250 from the city of Montes Claros in the State of Minas Gerais ) and 500 seronegative donors from the same time period , as previously described [19] . Recruited individuals underwent a health questionnaire , a medical evaluation , fasting blood sample collection for lipid profile , glucose analysis and NT-pro brain natriuretic peptide ( NT-proBNP ) , an ECG and an Echo . Results of ECG and Echo were reviewed blindly by centralized reading centers . T cruzi antibody status was confirmed on plasma samples collected at the time of enrollment . The study follows the Declaration of Helsinki of the Ethical Principles for Medical Research Involving Human Subjects , was approved by the Brazilian National Ethical Committee ( CONEP # 1312/2006 ) and all subjects gave written informed consent . In 1998–2002 , Fundação Pro-Sangue in São Paulo performed T . cruzi antibody screening using three serological methods: ELISA , hemaglutination and immunofluorescence for Chagas [20] . For the purpose of this study , we included as seropositive cases donors who were positive by all three assays at the time of donation and on a separate sample obtained at the time of counseling , generally one to four months after the donation . In Montes Claros , Hemominas screened all donations with two serological assays: ELISA and hemaglutination . All donors reactive to both assays at the time of donation and counseling were considered eligible for this study . Follow-up samples from all enrolled subjects were retested for T . cruzi antibody under code at the REDS-II Central Laboratory in the US using an FDA-approved assay manufactured by Ortho Diagnostics [21] . All 483 control donors tested negative on their follow-up samples collected at the time of enrollment and clinical assessments for this study . Of the 499 case donors who originally screened as seropositive at the time donation in 1998–2002 , serum collected in 2008–2010 from 498 tested repeat reactive , while the samples from one donor tested just below the assay cutoff consistent with slowly progressive seroreversion [21] . A face to face T . cruzi risk factor and health history questionnaire was administered by a trained nurse in each site . The questionnaire collected detailed information regarding demographics , cities of residence , physical activity , medical history , exposure to T . cruzi , previous ChD diagnoses , cardiac and GI symptoms , past medical history and medication history . All subjects received a physical examination by a non-blinded physician with recording of height , weight , blood pressure , heart rate and physical exam findings . Resting 12-lead ECG were recorded using the same model of machine at both sites ( General Electric MAC 1200 electrocardiograph; GE Healthcare , Waukesha , WI ) using standardized procedures . All ECGs were processed blindly by the central ECG laboratory ( Epidemiological Cardiology Research Center , Wake Forest University , Winston-Salem , NC ) , where they were visually inspected for technical errors and inadequate quality and processed with the 2001 version of the GE Marquette 12-SL program . ECGs were analyzed electronically , with manual over-reading by trained cardiologists to ensure quality control . ECGs were classified by Minnesota code criteria [18] using variables that were derived from the median complex of the Marquette measurement matrix . In this study , major and minor ECG abnormalities were defined as previously established [22] , modified to include ECG abnormalities typical of Chagas cardiomyopathy with prognostic significance , as frequent supraventricular or ventricular premature beats [10] . Old myocardial infarction ( MI ) on ECG was defined by the presence of major Q wave abnormalities ( MC 1 . 1 . x or 1 . 2 . x ) or minor Q waves abnormalities with ST segment or T-wave abnormalities ( 1 . 3 . x and [4 . 1 . x , 4 . 2 , 5 . 1 , or 5 . 2] ) [18] . Echocardiographic studies were performed using Sequoia 512 ultrasound machine ( Acuson , Mountain View , CA , USA ) at São Paulo site and GE Vivid3 ( GE Healthcare , Waukesha , WI ) at Montes Claros site . Cardiac measurements were performed according to the guidelines of the American Society of Echocardiography [23] . Studies were recorded in digital format and all measurements were performed on digital loops using a Digisonics offline analysis station ( version 3 . 2 software , Digisonics , Houston , Tex ) at the Cardiovascular Branch , Echocardiography Laboratory , National Heart , Lung , and Blood Institute , Bethesda , Maryland . LV ejection fraction ( LVEF ) was calculated based on modified form of Simpson's biplane method [23] and , if atrial fibrillation is present , LVEF is estimated by a visual method . For this study , LV systolic dysfunction was defined as LVEF<0 . 50 . Statistical analyses were conducted using SAS 9 . 2 and SPSS 18 . Distributions of data were examined for normality by using Kolmogorov-Smirnov tests . Continuous variables were expressed as median [interquartile range ( IQR ) ] and differences between seropositive and seronegative donors were compared using Wilcoxon-Mann-Whitney test , since these variables were not normally distributed . Categorical variables were summarized as counts and percentages and differences were compared using the Chi-square test or Fisher exact test . Association between quantitative ECG variables and LVEF was evaluated by Spearman correlation coefficient ( rs ) . A p-value<0 . 05 was considered significant . Receiver Operator Characteristic ( ROC ) curves were plotted in order to evaluate the accuracy of ECG measurements in detecting reduced LVEF and the area under the curve ( AUC ) was calculated . Sensibility , specificity and positive and negative predictive values of abnormal ECG , wide QRS duration ( ≥120 ms ) and long QTc interval ( >440 ms ) were calculated with 95% confidence intervals . These cut points were selected since they are well-established in cardiology practice .
The study sample consisted of 499 seropositive and 488 seronegative donors . Seropositive had a higher proportion of non-white skin color and lower weight , body mass index , total cholesterol levels and pulse heart rate , as well as higher NT-proBNP values ( table 1 ) . LV systolic dysfunction was more commonly found in seropositive cases . The groups were comparable with regard to age , gender or other major cardiovascular risk factors ( table 1 ) . Quantitative variables are shown in table 2 . All ECG intervals were longer in seropositive donors , and both heart rate and HRV indexes had lower values . ECG abnormalities were similar in seropositive and seronegative donors , except for a higher frequency in seropositive subjects of right bundle branch block and left anterior fascicular block , isolated ( 16% vs . 2% , p<0 . 001 for RBBB and 15% vs . 2% , p<0 . 001 for LAFB ) or in association ( 4% vs . 0 , p<0 . 001 ) ; and a higher frequency of left ventricular hypertrophy in seronegative subjects ( <1% vs . 1% , p = 0 . 004 ) . Seropositive cases showed more abnormal ECGs ( 51% vs . 32% , p<0 . 001 ) than seronegative donors due to a higher prevalence of major ( 26% vs . 9% ) ECG abnormalities . They also presented a higher number of major ECG abnormalities per tracing when compared to seronegative donors: 20% vs . 7% had one and 6% vs . 2% had 2 or more ECG abnormalities , respectively ( p<0 . 001 ) . LVEF was reduced in 36 out of 497 seropositive subjects with available data ( prevalence of LV systolic dysfunction of 7 . 2% ) ; in two patients LVEF measurement was not obtained due to technical reasons . Most seropositive subjects had LV systolic dysfunction considered mild ( LVEF ranging between 40 and 49% , n = 17 , 3 . 4% ) or moderate ( LVEF from 30–39% , n = 11 , 2 . 2% of total ) ; only 8 ( 1 . 6% ) showed markedly depresses LVEF ( <30% ) . Seropositive blood donors with and without LV dysfunction had comparable demographic and medical characteristics , although NT-proBNP levels were higher in those with LVEF below 50% ( 45 [25–76] vs351 [109–789] , p<0 . 001 ) . Seropositive donors with LV dysfunction showed longer PR , QRS and corrected QT intervals/durations ( Figure 1 ) , although both the heart rate and HRV indexes were not different between groups ( data not shown ) . Both QRS and QTc duration were associated with LVEF values ( rs: −0 . 159 , p<0 . 0003 , rs: −0 . 142 , p: 0 . 002 ) , and showed moderate accuracy in the detection of reduced LVEF ( ROC AUC: 0 . 778 and 0 . 790 , both p<0 . 0001 , Figure 2 ) . None of the other quantitative ECG variables showed significant correlation with measured LVEF . Several ECG abnormalities were more commonly found in seropositive donors with depressed LVEF ( table 3 ) , including rhythm disorders ( frequent supraventricular ectopic beats , atrial fibrillation or flutter and pacemaker ) , intraventricular blocks ( right bundle branch block and left anterior fascicular block ) and ischemic abnormalities ( old myocardial infarction and major and minor ST abnormalities ) . Almost all seropositive donors ( 33/36 ) with depressed LVEF showed at least one major or minor ECG abnormality . Only three subjects presented with normal ECG and abnormal LVEF: 0 . 46 . 0 . 45 and 0 . 40 , respectively; this last subject is a hypertensive patient with typical features of hypertensive cardiomyopathy in the Echo study who was not classified as having Chagas cardiomyopathy by the expert adjudication process employed in the parent study [19] . Seropositive donors with depressed LVEF had a higher prevalence of ECG abnormalities ( 69% vs . 23% , p<0 . 001 ) and a higher number of major ECG abnormalities per tracing when compared to seropositive donors with normal LVEF ( 39% vs . 19% with one and 31% vs . 4% with 2 or more abnormalities , p<0 . 001 ) The diagnostic performance of selected ECG abnormalities was evaluated ( table 4 ) . An abnormal ECG ( with minor or major abnormalities ) is a sensitive marker for recognition of seropositive donors with depressed LVEF , with a negative predictive value of 99% ( 96–100% ) .
In this analysis of ECG findings from a large , controlled study focused on former blood donors with and without T . cruzi seropositivity , the frequency of major ECG abnormalities was higher in seropositive donors than in seronegative subjects; right bundle branch block and/or left anterior were more commonly found in seropositive than in well-matched seronegative donors . When considering only seropositive donors , those with LV dysfunction rarely presented with a normal ECG , having one or more major ECG abnormalities , including rhythm disturbances , intraventricular blocks and ischemic abnormalities . The presence of either major or minor ECG abnormalities is therefore a sensitive marker of the presence of LV dysfunction in ChD , and , equally important , the absence of ECG abnormalities has high negative predictive value . Consequently , we recommend that initial clinical evaluation of seropositive blood donors ( and probably seropositive subjects identified through other population based screening programs ) can be limited to ECG with more expensive echocardiography performed on patients with abnormal ECG and/or clinical findings suggesting of ChD . Our findings suggest that seropositive blood donors have a similar profile to community ChD populations in terms of the prevalence and type of ECG abnormalities [12] , [14] , [15] , [24] , albeit with significantly lower rates and severity of abnormalities than in hospital and clinic-based series [11] , [16] , [16] , [25]–[28] . However , the mean age of blood donors was higher than those studied in early community-based series and the prevalence of concurrent chronic conditions in blood donors was relatively high ( 24% hypertension , 5% diabetes ) , in contrast to some studies in which those conditions were excluded [16] , [28] . Considering this higher frequency of ECG abnormalities in seronegative donors , right bundle branch block and/or left anterior fascicular block was the only finding more frequently found in seropositive subjects in relation to well-matched seronegative controls . This was also observed by Williams-Blandero et al . [14] , in one of the most recent of those community-based studies , with a similar age profile . Indeed , since the interruption of vector-mediated transmission has been achieved in many Latin American countries [2] , the age of T . cruzi infected subjects is increasing , and ChD is now a public health problem among older individuals in previously endemic regions [29] . LV systolic dysfunction , defined as reduced LVEF ( <0 . 50 ) , is a major marker of higher risk of death in ChD [30] . LV dysfunction , generally mild or moderate , occurred in a minority of cases in this sample ( 7 . 2% ) , reflecting the low risk profile of the seropositive donor population studied ( the donors had to be clinically asymptomatic to give blood donations in 1998–2002 , 8–12 years prior to the rigorous assessment including ECG and Echo that was the basis of the current analysis ) . In contrast , Ribeiro et al . [31] found a prevalence of more severe LV dysfunction ( defined as LVEF≤0 . 40 ) of 9 . 1% and 14 . 5% in two samples from a Brazilian Outpatient Clinic that is a regional reference center for blood banks and primary care units , and Salles et al . [28] observed LV dysfunction in 109 out of 738 patients ( 14 . 8% ) at another Brazilian Reference Outpatient Clinic . In this study , several ECG abnormalities typical of ChD were predictive for LV dysfunction among seropositive donors . Right bundle branch block , frequently combined with left anterior fascicular block , is the most characteristic ECG abnormality in ChD [10] and is associated with higher risk of death in longitudinal studies [13] , [32] . ChD patients with pacemakers have lower LVEF comparative to pacemaker patients without ChD [33] . In contrast to what we found in this study , frequent ventricular ectopic beats have been repeatedly related to LVEF depression in ChD [26] , [34] and the observed association of supraventricular ectopic beats with LV dysfunction has not been reported before . Since frequent supraventricular ectopic beats can precede the development of atrial fibrillation , we hypothesize that higher left atrial pressure and volume secondary to LV systolic dysfunction may lead to frequent atrial ectopic beats and , after years , to atrial fibrillation . Atrial fibrillation is a late abnormality in the natural history of ChD , related to other ECG abnormalities , LV dysfunction and higher risk of death [12] , [35] . Pathological q waves , ischemic ST-T abnormalities and abnormal T waves have also been reported to be markers of risk in Chagas cardiopathy [27] , [36] . Prolonged QRS duration and QTc interval were both related to LV dysfunction [27] , [37] and to worse prognosis [27] , [38] . Since prolonged excitation time in ventricular conduction defects may induce secondary prolongation of the QT interval [39] , the significance of prolonged QTc interval in seropositive donors should be interpreted with caution , considering that no correction for QRS duration was made in this study . Those with more than one ECG abnormality are at greater risk of having LV dysfunction , as reported in other studies [26] , [40] . Heart rate variability indexes SDNN and RMSSD , calculated from standard 10-seconds ECG tracing , were reduced in seropositive donors but these findings were not correlated to the presence of LV systolic dysfunction . Both SDNN and RMSSD indexes from a 10-second ECG are markers of parasympathetic modulation of the sinus node [41] . In some studies in other settings [42]–[44] , 10-s HRV indexes were predictive of the risk of death . Impairment of cardiac vagal modulation has been consistently reported in ChD [45]–[47] and occurs early in the evolution of the disease , preceding LV systolic dysfunction [48] . However , the association of vagal dysautonomia with LVEF and with prognosis in ChD is still controversial [46] , [49] . Data from our study suggest that the ECG can be useful to guide the management of seropositive blood donors: an abnormal ECG is a sensitive marker of LV dysfunction , while a normal ECG carries a high negative predictive value . Indeed , a normal ECG is an established marker of excellent prognosis in medium-term follow-up of T . cruzi seropositive subjects [12] , [13] , [16] . The diagnostic performance of a single ECG interval measurement , such as a normal QRS duration or corrected QT interval , is not good , as previously reported [37] . The main strength of this analysis is that it is based on findings from a large controlled , rigorously conducted study , with central and blinded reading of both ECG and Echo results , and with classification using an internationally accepted ECG code , the Minnesota Code . This is , to the best of our knowledge , the only study with these features in a population of seropositive and matched seronegative blood donors . Because these blood donors were healthy at study baseline , the incidence of pathology will be lower than in prevalent patient cohorts . On the other hand , because patients with common chronic conditions , including hypertension and diabetes , were not excluded , it allows the generalization of findings to other seropositive blood donor populations in endemic and non-endemic countries . The main limitations are related to the cross-sectional status of the current analysis , with no information on the prognostic role of the observed abnormalities . We also lacked baseline measurements on the cohort to assure the absence of cardiac pathology , although all were healthy enough to donate blood . Moreover , blood donors from non-endemic countries may have different epidemiological profiles due to country of origin , socio-economic status and time since acquisition of infection prior to detection as seropositive donors , as well as possible differences in disease progression depending on the strain of T . cruzi . In conclusion , we identified several ECG abnormalities that are predictive of LV dysfunction in ChD . Due to the study setting involving previously healthy seropositive donors who developed incident ChD , these study findings may be extrapolated to other low-risk populations . In particular , the results may guide the evaluations of patients with incidentally detected T . cruzi seropositivity from blood bank testing in endemic and increasingly in non-endemic countries , and from public health screening in endemic countries . | Chagas disease ( ChD ) , caused by the protozoa Trypanosoma cruzi , is endemic in most Latin America countries and may be transmitted via blood transfusions . Cardiac disease is a major feature of chronically infected patients and may be lethal . Universal blood bank screening for ChD has been established in most Latin American countries , as well as in non-endemic countries with large immigrant populations , including the United States , Canada , Spain and Portugal . Blood donor screening leads to large numbers of new diagnoses of chronic T . cruzi infection . Counseling these individuals should address the recognition of those with more severe disease that deserve to be rigorously evaluated by experienced cardiologists and treated more promptly . The electrocardiogram is an important exam that can help in the recognition of cardiac disease and the evaluation of prognosis in ChD patients , but its role in blood donors has not been studied . The authors describe some electrocardiographic abnormalities that are typical of the infected blood donors , as well ECG abnormalities that help in the identification of those with severe cardiac involvement . These results may guide the evaluations of patients with incidentally detected T . cruzi infection from blood bank testing or public health screening . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"chagas",
"disease",
"neglected",
"tropical",
"diseases"
] | 2013 | Electrocardiographic Abnormalities in Trypanosoma cruzi Seropositive and Seronegative Former Blood Donors |
Aedes aegypti is a primary vector of dengue , chikungunya , Zika , and urban yellow fever viruses . Indoor , ultra low volume ( ULV ) space spraying with pyrethroid insecticides is the main approach used for Ae . aegypti emergency control in many countries . Given the widespread use of this method , the lack of large-scale experiments or detailed evaluations of municipal spray programs is problematic . Two experimental evaluations of non-residual , indoor ULV pyrethroid spraying were conducted in Iquitos , Peru . In each , a central sprayed sector was surrounded by an unsprayed buffer sector . In 2013 , spray and buffer sectors included 398 and 765 houses , respectively . Spraying reduced the mean number of adults captured per house by ~83 percent relative to the pre-spray baseline survey . In the 2014 experiment , sprayed and buffer sectors included 1 , 117 and 1 , 049 houses , respectively . Here , the sprayed sector’s number of adults per house was reduced ~64 percent relative to baseline . Parity surveys in the sprayed sector during the 2014 spray period indicated an increase in the proportion of very young females . We also evaluated impacts of a 2014 citywide spray program by the local Ministry of Health , which reduced adult populations by ~60 percent . In all cases , adult densities returned to near-baseline levels within one month . Our results demonstrate that densities of adult Ae . aegypti can be reduced by experimental and municipal spraying programs . The finding that adult densities return to approximately pre-spray densities in less than a month is similar to results from previous , smaller scale experiments . Our results demonstrate that ULV spraying is best viewed as having a short-term entomological effect . The epidemiological impact of ULV spraying will need evaluation in future trials that measure capacity of insecticide spraying to reduce human infection or disease .
Aedes aegypti is a primary vector for dengue ( DENV ) , chikungunya ( CHIKV ) , Zika ( ZIKV ) and urban yellow fever viruses ( YFV ) . Dengue has become the most important human arthropod-borne viral infection worldwide [1 , 2] . Each of these pathogens can be associated with explosive epidemics , where high disease incidence and public fear combine to overwhelm health systems [3] . Such epidemics put intense pressure on public health departments to react with emergency vector control measures [4 , 5] . Ae . aegypti adults are primarily diurnal and females take frequent blood meals , predominantly from humans [6–8] . These behaviors can in part explain why Ae . aegypti has been associated with epidemic virus transmission even when its population densities are low [9] . Because adults typically reside inside houses [8] where food , mates , and oviposition substrates are readily available , indoor adulticide space spraying has been more effective than outdoor spraying for suppressing Ae . aegypti populations in small scale evaluations [4 , 10 , 11] . When indoor space sprays are applied appropriately , in carefully controlled small-scale experiments , adult Ae . aegypti populations often decreased by >80% . Population densities typically recovered quickly , however , [12–15] due to emergence of nulliparous mosquitoes from larval aquatic habitats inside sprayed areas [11] , through migration from locations outside of sprayed areas [15] , or from females in sprayed houses that survived . In a systematic literature review , Esu et al . [4] found only six studies from 1970’s to 2010 that tested ultra-low volume ( ULV ) indoor space spraying under natural field conditions that met minimum standards for evaluating mosquito population suppression . None of the studies evaluated the impact of these methods on human virus infection or disease [4] . Results ranged from immediate reduction in biting by 99% and adult population reduction lasting six months [16] , to a more common , modest control lasting 1–5 weeks [14 , 15 , 17] . Most studies were small scale , with each treatment typically including one replicate of less than 50 houses . It is important to distinguish between indoor and outdoor space spraying , the latter usually delivered by vehicle-mounted machines and whose effectiveness is often limited by a failure of insecticide droplets to penetrate indoors where Ae . aegypti rest [11] . A more recent review of vector control effectiveness for dengue [18] concluded that “although space spraying is the standard public health response to a dengue outbreak worldwide , and is recommended by WHO [19] for this purpose , there is scant evidence available from studies to evaluate this method sufficiently . ” In fact , Bowman et al . [18] could find no well-designed trial that assessed the impact of non-residual indoor or outdoor space spraying on human dengue infection or disease . Ae . aegypti populations in the Amazonian city of Iquitos , Peru have been studied extensively since 1998 . The spatial distribution of the species is highly clustered and does not have a consistent spatial or temporal structure [20 , 21] . Adult and immature population indices are highly variable and subject to sampling error [22] . Evaluation of control measures for this species , therefore , requires large sample sizes and exhaustive sampling . In addition to studying the mosquito itself , the Iquitos research program monitored DENV transmission through passive clinic-based febrile surveillance in health care facilities throughout the city [23] and a series of prospective cohort studies in targeted city neighborhoods [24–26] . The combination of longitudinal entomological and epidemiological studies created a database that could be used to examine , in real time , the impact of Ministry of Health ( MoH ) vector interventions on Ae . aegypti populations and human disease . During their interventions , the MoH sprayed non-residual insecticide inside homes three times over an approximately 3-week period [27] . Over a 10-year period , this kind of citywide municipal vector control program was associated with significant decreases in Ae . aegypti adult populations [28 , 29] and when interventions were applied during the first half of the dengue transmission season , fewer dengue cases were detected and the transmission season was shorter [27] . While the qualitative results from that analysis of dengue are consistent with an expectation of a positive public health impact of intra-domicile ULV insecticide application on dengue incidence , more statistically robust epidemiological studies are needed [30] . Prevention of Aedes-transmitted viral disease will require integrated approaches; i . e . , combinations of existing and/or novel vector control strategies as well as vaccination . Mathematical models provide a way to compare diverse strategies and identify the most promising approaches . For example , data on Ae . aegypti populations in Iquitos were used to develop a biologically detailed , spatially explicit , stochastic model that tracked Ae . aegypti dynamics and genetics in an 18-ha area of the city [31 , 32] . Preliminary validation of the model using Iquitos data was carried out [31] , but evaluation of its capacity to accurately predict the entomological outcome of a vector control perturbation had not been tested . The experiments described here were primarily designed to generate data that could be used to test the ability of the entomological model to predict impacts of suppression measures . In this study , we carried out a large-scale evaluation of the entomological impact of a widely used emergency vector intervention of Aedes-transmitted viruses in a well-characterized study site . Our specific goal was to evaluate the impact of 6 cycles of indoor ULV pyrethroid spray applications ( hereafter referred to as “spray applications” ) on reductions of Ae . aegypti populations . Our experiments spanned periods of relatively low and high Ae . aegypti density in Iquitos , and compared the ULV application in experimental and public health settings . Our results constitute an important data set for development and validation of Ae . aegypti population dynamics models , and provide a detailed account of indoor space spray effects on Ae . aegypti populations .
The study protocol was approved by the Naval Medical Research Unit Six ( Protocol #NAMRU6 . 2013 . 0001 ) Institutional Review Board , which included Peruvian representation , in compliance with all US Federal and Peruvian regulations governing the protection of human subjects . IRB authorization agreements were established between the Naval Medical Research Unit Six and the University of California at Davis and North Carolina State University . The protocol was reviewed and approved by the Loreto Regional Health Department , which oversees health research in Iquitos . In all instances consent from adult members of houses was obtained without written consent . Written information sheets were provided to study participants , providing a detailed overview of the experiment design , procedures , and study goals before initial pre-interventions surveys . Permission to enter houses was provided at each survey or spray application visit . The studies were conducted in two neighborhoods in the Maynas district of Iquitos ( Fig 1 , Maps ) . Iquitos has a human population of ~380 , 000 ( 73 . 2°W longitude , 3 . 7°S latitude , 120 m above sea level ) . Located in the Amazon Basin of northeastern Peru , Iquitos is the largest urban center in the Department of Loreto , and has an average daily temperature of 25°C and an average annual precipitation of 2 . 7 meters . Dynamics of Ae . aegypti populations in Iquitos are described in detail in earlier publications [20–22 , 24 , 33–38] . These neighborhoods were selected because , historically , they consistently had high Ae . aegypti densities [33] . In addition , the area was spatially configured to meet our study design of a central spray area surrounded by buffer area that would serve as a control . Both experimental study neighborhoods were characterized by city blocks of row houses ( dwellings that share walls ) . Most houses occupied lots that were narrow ( 3–10 m wide ) , but relatively deep ( 20–60 m long ) . The majority of houses served as family residences , often containing extended or multiple families . Some houses were used for small businesses or offices , and others were unoccupied . There were a small number of vacant lots containing no structures ( <1% ) . Many study houses were mixed-purpose , sharing living areas with a small store ( “bodega” ) , office , shop ( e . g . carpentry or vehicle repair ) , or restaurant . Vector control activities were ongoing in Iquitos . The MoH carried out regular entomological surveillance and larviciding activities with temephos ( Abate ) at ~3 month intervals . Since 2002 , with few exceptions , MoH carried out 1–3 emergency indoor pyrethroid spray campaigns per year in response to dengue outbreaks , with variable success [27] . Our study was completed in 2014 , while resistance bioassay profiles prior to January 2013 indicated Ae . aegypti populations in the city were susceptible to pyrethroids [39] . Fig 2 summarizes the design of our two separate experiments . The first and smaller of the two experiments ( S-2013 ) ran for 16 calendar weeks and included an experimental buffer sector that was not sprayed , surrounding a central experimental sector that was sprayed . The buffer sector contained 765 houses and the spray sector had 398 houses ( Fig 1A ) . The S-2013 study area was located on the western border of the city , proximal to Lake Moronacocha ( Fig 1C ) . The larger second experiment ( L-2014 ) ran for 44 calendar weeks , and included 1 , 051 houses in the surrounding buffer sector and 1 , 110 houses in the central spray sector ( Fig 1B ) . L-2014 was carried out in a neighborhood several kilometers to the north of S-2013 , centrally located in Iquitos , and bordered on the south by an abandoned airstrip ( Fig 1C ) . The L-2014 study area was selected because the Ae . aegypti-free airstrip provided a physical barrier to Ae . aegypti dispersal on one of its four sides . This experimental structure of L-2014 was selected to test our mathematical model’s ability to capture any spatial features of the recovering mosquito population . To monitor population densities and age structure of Ae . aegypti populations , standardized adult mosquito collections were carried out using Prokopack aspirators [40] ( henceforth adult surveys ) and standardized larval/pupal demographic surveys [41–43] ( henceforth immature surveys ) were undertaken , except when noted . Survey protocols are described in detail in previous publications [20–22 , 35 , 40] . Collected adults were immediately transported to a field laboratory in Iquitos for processing as described in Morrison et al . [33] . Adult mosquitoes were sedated by cold ( 4°C ) , identified , counted , and females separated . In most cases , female Ae . aegypti were scored as unfed , blood fed ( full , half full , or trace amounts ) , or gravid . Females were also scored for parity [44] . Both the S-2013 and L-2014 experiments consisted of 6 cycles of spray applications , which were applied over 6 weeks . Spraying was done by MoH employees between 17:00–20:00 to avoid high temperatures and varying winds . Each spray team was comprised of 3 individuals: 2 MoH sprayers and 1 monitor from the research team . Each week , on the initial day of a spray cycle ( usually Mondays ) , spraying was attempted in all houses in the spray sector . To improve spray coverage within each cycle , on subsequent days spray teams revisited houses that were not sprayed on the initial day of the spray cycle ( a minimum of 2 and up to 10 visits , as needed ) to conduct spraying . Pyrethroid insecticides were applied using Solo or Stihl backpack sprayers with settings adjusted for ULV application , or Colt hand-held ULV sprayers . Residents were instructed not to return to their houses for a minimum of 1 hour . See S1 Text for more details . A MoH emergency intervention interrupted the L-2014 experiment with a set of spray applications ( 3 cycles over 2 weeks ) that were applied similarly to the above experimental sprays . The principal differences were as follows: during the emergency intervention , MoH personnel generally sent an advance team with loudspeakers announcing the arrival of the spray teams; MoH personnel visited each house on a block a single time ( they had no mechanism to spray houses missed on their initial visit ) ; and pyrethroids were applied to both the spray and buffer sectors using only Solo or Stihl backpack aspirators . As a quality control measure , for each spray cycle , 3 to 7 houses were selected to monitor efficacy of the insecticide spray . Operators did not know which houses would be selected for monitoring . For each monitored house , just after the spray operator had finished the application , a single screen cage containing adult mosquitoes was placed in each of the following locations: bedroom , living room , kitchen , and yard , based on standard WHO protocols [45–47] . Each cage contained 25 adult Ae . aegypti of age 24–36 hours from a pathogen-free laboratory colony [45 , 46] . A separate laboratory colony was initiated for each experiment from mosquitoes collected from houses in Iquitos and held for 1–2 generations prior to use . One hour after spraying , all cages were retrieved and evaluated for knockdown ( no movement ) , stored in a styrofoam cooler with moist paper towels for 24 hours , and then examined for mortality . When mortality was <80% , equipment was recalibrated to ensure proper spray function on subsequent days . Experimental study sectors are depicted in Fig 1A & Fig 1B . We refer to the temporal sampling units as “circuits” because they were time periods when full survey routes through all of the blocks of houses in the spray and buffer sectors were completed ( see Fig 2 for a flow chart of experimental design , and S7 Fig for survey maps ) . During each circuit , attempts were made to visit and survey 100% of the houses in the entire study area at least once ( with one exception , L-2014 C2 ) . The percentage of total houses successfully surveyed and/or sprayed in each circuit ranged from 67–90% , due to closed or unoccupied houses , or residents who chose not to participate in the study ( see Fig 3B , S1 Table ) . Each circuit was divided into subcircuits that lasted approximately one week , but never more than 10 days . In general , subcircuit surveying was conducted systematically by block , with surveyors attempting to visit every 4th house ( 25% of the circuit ) each week ( see S2 Text for exceptions ) . Both experiments consisted of 6 weekly cycles of ULV indoor spray applications ( see above ) . Immature and adult surveys were carried out before ( pre-intervention ) and after ( post-intervention ) the spraying periods . During the experimental spray periods , only adult surveys were carried out ( no immature surveys ) . In the baseline pre-intervention circuit of each experiment ( C1 ) , 8 study teams surveyed the entire study area ( 2 people / team ) , proceeding by block until all houses in the study area were visited at least once ( see below for details regarding L-2014 ) . Houses that were not accessible on a day of a visit were revisited the next day and surveyed if open . After all study blocks were surveyed , houses that remained unsurveyed were visited a final time , and surveyed if possible . In subsequent circuits , similar spatially systematic surveying within subcircuits was carried out , and unsurveyed houses were visited a minimum of 3 times per circuit , or until access was obtained or refused . The initial S-2013 baseline pre-intervention circuit ( C1 ) was carried out from 22–29 April 2013 in the spray sector , and from 29 April-16 May 2013 in the buffer sector ( C1 , Fig 3A ) . During the experimental treatment circuit ( C2 ) , Alphacypermetrin 10% ( Turbine 10% ) was applied once per week for 6 consecutive weeks using Solo backpack sprayers ( Cycles 1–6 ) or Colt hand-held sprayers ( Cycles 4–6 ) . Adult surveys were typically carried out during the spray period on Monday afternoons just prior to the initiation of each spray cycle , as described above . This design , therefore , measured adult densities up to 7 days after a previous spraying event . Post-intervention surveys ( C3-C4 ) were initiated 10 days after completion of the last spray cycle ( see S7A Fig and S8A Fig for detailed maps of surveys and sprays , respectively ) . In contrast to S-2013 , the L-2014 study teams worked in two groups during the pre-intervention baseline circuit ( C1 ) , with one group in each sector so as to survey both sectors simultaneously . Each group consisted of 4 two-person teams . Following the initial L-2014 baseline , pre-intervention circuit ( C1 ) , the experiment was interrupted by a MoH citywide emergency intervention in response to a dengue outbreak ( see also S1 Text ) . Unless otherwise noted , only Ae . aegypti data were analyzed , and houses were used as the basic spatial units of observation . A "spray status" indicator variable was assigned to each house survey conducted in the spray sector during experimental spray periods . "Prior spray" indicated that a spray application occurred in a surveyed house prior to the survey during the current or previous calendar week ( otherwise , "no prior spray" ) . During L-2014 , the relative timing between spray and subsequent survey was unclear for a limited number of surveys , which were designated as "timing unclear" ( S4 and S5 Tables ) . For each experiment , a suite of statistical models was developed to estimate the impact of spray treatment on mosquito densities , proportion of infested houses , and population age structure ( as determined from parity examination ) . With one exception , all comparisons and significance tests were conducted within-experiment . Two generalized linear model ( GLM ) specifications were employed , both of which used a log link . For all counts , a negative binomial GLM ( NB-GLM ) was used . Here , the response was the count of mosquitoes per house , and was assumed to follow a negative binomial distribution . The NB-GLM estimates the log of mean counts , and is akin to Poisson regression , while allowing for response over-dispersion ( separate mean and variance ) [48] . For all proportions , a logistic GLM ( L-GLM , i . e . , logistic regression ) was used . Here , the response was the proportion of successes ( out of total number of events ) , and was assumed to follow a binomial distribution . The choice of “success” was an arbitrary label applied to one of two mutually exclusive possibilities ( presence or absence ) . The L-GLM estimates the log probability of success . For ease of interpretation , all model results were un-transformed after analysis and displayed in the original ( unlogged ) scale of observations . To identify structural , pre-perturbation differences between sectors , an NB-GLM was used . This estimated the number of Ae . aegypti adults per house ( AA/HSE ) in the baseline circuit ( C1 ) in response to physical characteristics of houses , including building , floor , and roof construction , as well as number of containers , rooms , and surveyed rooms . To assess the effect of spraying , an NB-GLM was employed to estimate AA/HSE in response to circuit and spray sector . In addition , a companion L-GLM was used to predict Adult House Index ( AHI: proportion of houses with 1 or more Ae . aegypti adults ) in response to circuit and spray sector . Finally , the NB-GLM model formulation was tested with alternate responses: female Ae . aegypti adults per house , and non-Aedes adults per house . An NB-GLM was also used to estimate the effect of study year and spray status on AA/HSE . This model included only surveys conducted in the spray sector during experimental spray periods . Counts from immature surveys and parity surveys were converted to proportions: container surveys yielded per-house proportion of positive containers ( henceforth called the PrPC ) , which is also referred to as the container index . Parity surveys yielded the per-house proportion of nulliparous females ( henceforth called the PrNF ) . Each proportional measure ( PrPC , PrNF , and PrIH ) was analyzed using a pair of L-GLM , weighted by the number of observations , with a separate model for each study year . Predictors included circuit and sector . The response was the log proportion of “successful” events per house , i . e . , detection of positive containers or nulliparous females . The container model estimated the log proportion positive containers per house , log ( PrPC ) , and the reproductive status model estimated log proportion nulliparous females per house , log ( PrNF ) . The total number of Ae . aegypti positive containers per house ( PC/HSE ) was modeled using an NB-GLM . Note that Breteau Index ( BI ) = 100* ( PC/HSE ) . To further evaluate the effect of spraying on mosquito densities , contrast analysis [49] was employed on the sector-by-circuit NB-GLM . Contrasts were made between circuits ( spray sector only ) , and between sectors . The between-circuit contrast was complicated by temporal variation , either in extrinsic environmental factors , such as weather , or in intrinsic ecological processes , such as demographic stochasticity . The between-sector contrast was complicated by potential spatial ecological differences between sectors . More robust conclusions can be made if both types of contrasts provide similar assessments of the effect of spraying . For the statistical models of adult , immature , and parity surveys , statistically indistinguishable groups and 95% confidence intervals ( CI ) of experimental group effects were estimated using least-squares means , also known as predicted marginal means , via the lsmeans R package [49] . Tukey's method was used to control the family-wise error rate [49] .
During S-2013 , 1 , 860 ULV spray applications ( 6 weekly cycles ) were carried out in 398 houses . During L-2014 , 4 , 986 spray applications ( 6 weekly cycles ) were carried out in 1 , 110 houses . During the MoH emergency spray campaign that interrupted the L-2014 experiment there were 3 , 502 spray applications in 1 , 617 houses over 10 calendar days . A total of 3 , 843 surveys over 16 weeks and 12 , 124 surveys over 44 weeks were carried out in S-2013 and L-2014 ( including MoH emergency campaign ) , respectively ( Fig 3A , Table 1 ) . Adult Ae . aegypti densities were highly variable over space ( S1 Fig ) and time ( S4 Fig ) with highly skewed distributions . No adult mosquitoes were collected from most houses , and large numbers of adults were captured in very few houses ( S1 Fig ) . Model results ( AA/HSE and AHI ) are shown in Fig 4; model contrasts ( AA/HSE ) are shown in Fig 5; details of adult densities and house indices are shown in S6 and S7 Tables . Overall , adult densities in the S-2013 baseline circuit ( early May , C1 ) were 0 . 26 and 0 . 40 Ae . aegypti per house ( AA/HSE ) in the buffer and spray sectors respectively . During this same baseline circuit , 15% and 16% of houses contained one or more Ae . aegypti adults ( AHI ) in the buffer and spray sectors , respectively ( S7A and S7B Table ) . The L-2014 baseline circuit ( January , C1 ) showed that Ae . aegypti adult densities were higher than in S-2013: 0 . 62 and 0 . 77 AA/HSE in the buffer and spray sectors , respectively . A later pre-intervention circuit in April ( C5 , prior to experimental spraying ) yielded 0 . 44 and 0 . 67 AA/HSE in the buffer and spray sectors , respectively . The corresponding AHIs for these surveys were 31% and 34% in the spray and buffer sectors , respectively for January , C1 , and 22% and 28% for April , C5 . Adult Ae . aegypti densities and house indices within the spray sector during spray periods were also lower during S-2013 ( 0 . 07 AA/HSE; AHI 5 . 5% ) compared to L-2014 ( emergency spraying , C3: 0 . 30 AA/HSE; AHI 18%; experimental spraying , C6: 0 . 31 AA/HSE; AHI 11% ) . In the S-2013 post-intervention circuits ( C3-C4 ) , Ae . aegypti adult densities in the spray sector achieved a maximum of 0 . 35 AA/HSE ( AHI 23% ) . In L-2014 ( C7-C9 ) , Ae . aegypti adult densities in the spray sector reached a maximum of 1 . 31 AA/HSE ( AHI 41% ) . Meteorological conditions were consistent between the two experiments , with average temperatures of 25 . 5°C ( average minimum = 22 . 0°C , average maximum = 32 . 0°C ) and 25 . 6°C ( average minimum = 22 . 0°C , average maximum = 31 . 9°C ) during the S-2013 and L-2014 experiments , respectively ( National Climatic Data Center , https://www . ncdc . noaa . gov/cdo-web/ ) . Precipitation during both years was approximately 0 . 84 cm per day . During the L-2014 entomological surveys for the MoH emergency citywide spray operation ( January-March 2014 ) , the temperatures were higher ( average 25 . 9°C , average minimum = 23 . 3°C , average maximum = 32 . 6°C ) and it was rainier ( average 1 . 09 cm per day ) than at other times during the S-2013 and L-2014 experiments . Comparisons of spray and buffer sectors in both experiments indicated that the two sectors had similar housing characteristics . No household physical characteristic was a predictor of adult mosquito density . Consequently , these characteristics were not included in our statistical models . Overall , for both years baseline numbers of Ae . aegypti adults were comparable between spray and buffer sectors ( S2 Table ) . During S-2013 , however , a marginally significant difference was found between the buffer and spray sectors during the baseline ( C1 ) circuit ( 0 . 26 vs 0 . 40 AA/HSE , resp . ; Fig 5 , S2 Table , p = 0 . 039 ) , making some statistical analyses of spray impacts conservative . During L-2014 , baseline densities ( C1 ) were not significantly different between the buffer sector ( 0 . 62 AA/HSE , AHI = 31 . 1% ) and spray sector ( 0 . 77 AA/HSE , AHI = 33 . 7% ) ( Fig 5 , S2 Table , p = 0 . 09 ) . No statistically significant baseline differences in adult female age structure were observed between buffer and spray sectors ( PrNF , S8 Table ) . Baseline immature indices were similarly not different; for example , Breteau Indices ( BI = 100 * PC/HSE ) ranged from 9 . 4–10 in the buffer and spray sectors in both experiments ( S9 Table ) . Container indices ( i . e . , percentage of water-holding containers infested with larvae or pupae , 100*PrPC ) ranged from 3 . 9–4 . 1 in S-2013 and 3 . 1–3 . 3 in L-2014 ( S10 Table ) . The average percent of houses sprayed was lowest during the 3 MoH citywide emergency spray cycles in L-2014 , ranging from 71% during cycle 1 to 62% in cycle 3 ( Fig 3B ) . For S-2013 , coverage started at 77% in cycle 1 , decreased to 73% in cycle 3 , and then improving in each subsequent cycle to 90% ( cycle 6 ) . During L-2014 experimental spraying , coverage started at 74% in cycle 1 , then modestly increased over time to approximately 82% in cycle 6 ( Fig 3B ) . In both experiments , most spray sector houses were sprayed in more than 3 out of 6 spray cycles , and more than half of target houses were sprayed in all 6 spray cycles ( S2 Fig ) . The primary reasons for not spraying a house were: house closed when personnel visited ( 3–16% for S-2013 spray , 19–28% for L-2014 MoH emergency spray , 7–16% for L-2014 spray ) , or residents did not allow access to the house ( 6–14% for S-2013 spray , 9–11% for emergency spray , 8–11% for L-2014 spray ) . During the S-2013 experiment , but not in L-2014 , the reasons given by residents for refusing access were recorded . In many cases , teams were allowed access on subsequent visits . In early cycles , about one-third of the refusals cited a direct objection to fumigation , saying they did not believe it was effective or that the teams were not really using insecticide . In other cases , the reason given was inconvenience to the residents: eating , bathing , working , selling food , or that a sick person or newborn was in the house and could not leave . In some instances , the homeowner was not present and consent could not be given . During S-2013 , 24-hour mortality of caged sentinel mosquitoes ranged from 87–97% with some variation across cycles ( S3 Fig ) . Mean mortality was lower in L-2014 , ranging from 53–87% . Overall , a significant decrease in spray efficacy was observed in L-2014 relative to S-2013 ( Table 2 ) . During S-2013 , Colt hand-held ULV sprayers were used on 1/3rd of the blocks during spray cycles 4–6 . Higher mortality and knockdown were observed in cycles 4–6 , while less variation was observed in cycles 1–3 , which employed only backpack sprayers . Droplet size ( mean±SD ) varied between experiments and sprayer type . Colt sprayers had smaller and more consistent droplets ( 19 . 1±12 . 6 μm ) than backpack sprayers ( 29 . 2±19 . 5 μm ) . During the L-2014 MoH emergency spray , backpack sprayers were not properly calibrated during the initial cycle , with an average droplet size of 39 . 8±25 . 8 μm . This improved to 20 . 6±14 . 1 μm in subsequent cycles . During the L-2014 6-cycle experiment , droplet size averaged 18 . 1±14 . 7 μm and 23 . 6±13 . 2 μm for Colt and backpack sprayers , respectively . Surveys conducted during the 6-week spray period ( C2 ) generally occurred about one week after spraying . During the spray period , ULV spraying reduced adult Ae . aegypti population densities rapidly and significantly from 0 . 40 to 0 . 07 AA/HSE after six cycles of spraying ( Fig 4 ) , yielding an 82 . 5% reduction relative to baseline ( Fig 5 , S3 Table , p<0 . 00001 ) . The buffer sector , in contrast , had 0 . 26 AA/HSE both before ( C1 ) and during ( C2 ) the spray period . Adult densities in the sprayed sector were 73 . 1% lower than in the buffer sector during the spray period ( C2 , Fig 5 , S2 Table , p<0 . 00001 ) . Ongoing surveys within the spray sector during the spray period ranged from 0 . 04–0 . 08 AA/HSE , and did not change significantly over the course of the six sprays ( Fig 6 ) . Spray sector AA/HSE remained 45% lower than baseline levels during the first post-intervention period ( C3 , Fig 5 , S3 Table , p = 0 . 035 ) , but densities increased from 0 . 04 to 0 . 27 AA/HSE between the first and second week post-spray . During the second post-intervention period ( C4 ) , spray sector adult densities returned close to baseline densities , increasing from 0 . 22 to 0 . 35 AA/HSE ( S6A Table ) which was 89% of baseline ( Fig 5 , S3 Table , p = 0 . 94 ) and 83 . 3% of the buffer sector density at that time ( Fig 5 , S2 Table , p = 0 . 36 ) . Adult house indices in the spray sector , by comparison , decreased from 16% during baseline surveys to 5 . 5% during the spray period ( C2 ) , then increased to 12 . 7% and 17 . 3% during the first and second post-intervention periods , respectively ( C3-C4 , S7A Table ) . In the buffer sector , AHIs were 15% during both baseline and spray periods , then increased to 21% and 23% in the first and second post-intervention evaluations ( S7A Table ) . During the S-2013 spray period ( C2 ) , only a small number of females ( 9 total ) were collected in the spray sector ( S8A Table ) . Therefore , no attempt was made to compare the age structure of Ae . aegypti populations before and after spray applications for this experiment . Model estimates of the proportion of nulliparous females ( PrNF ) showed accordingly high uncertainty ( S5B Fig and S8A Table ) . Results from pupal demographic surveys followed a pattern similar to that of adult house indices . Baseline BIs were 10 . 0 in both the buffer and control sectors ( S9A Table ) . BIs were not measured during the spray period; during the first post-intervention period ( C3 ) , however , BI decreased slightly in the spray sector to 7 . 4 and increased to 16 . 1 in the buffer sector . During the second post-intervention period ( C4 ) , BIs were 15 . 1 and 22 . 3 in the buffer and spray sector , respectively . The post-treatment spray sector had statistically significantly higher PrPC than any other sector or time period ( S5B Fig , S10A Table ) . In the spray sector during C4 , PrPC reached approximately 0 . 11 , significantly higher than seen in the spray sector during the baseline C1 ( 0 . 04 ) or in the buffer sector during C4 ( 0 . 06 ) . During the S-2013 experiment , entomological surveys were carried out during the afternoon before each ULV spray cycle was initiated . For the majority of S-2013 house surveys with a prior spray application , Ae . aegypti densities were measured 7 days after the previous spraying , and 308 out of 311 surveys with prior sprays occurred from 5 to 8 days after the spray application . In contrast , L-2014 house surveys with prior sprays were typically conducted 1 to 4 days after the surveyed house's prior spray . This difference was the result of logistical concerns , as L-2014 involved many more houses . For 164 of the 1 , 054 house surveys with prior sprays during L-2014 ( 16% ) , the exact timing of the spray event relative to the subsequent survey was not available . In addition , some houses within the spray sector were surveyed without a prior spray ( S5 Table ) . Thus , the average interval between a house's spray application and subsequent survey was shorter in L-2014 than in S-2013 ( median 2 days and 7 days , respectively ) . In both experiments , AA/HSE were lower in spray sector houses that had been sprayed prior to surveying compared to those that had not . In S-2013 , AA/HSE was 0 . 06 and 0 . 11 in houses with prior spray and no prior spray , respectively , while L-2014 experienced 0 . 28 and 0 . 56 AA/HSE in houses with prior spray and no prior spray , respectively ( S5 Table ) . A ( marginally ) significant difference in AA/HSE between spray status groups was observed only during 2014 ( S4 Table , p = 0 . 047 ) .
Despite the lack of a well-informed evidence base [18] , vector control of Ae . aegypti is often described as ineffective yet continues to be widely practiced by public health programs [11 , 18 , 50–52] . Increasing attention has been given to integrated vector management , community involvement , and sustainability [53] . There is increasing recognition , however , that programs lacking interventions specifically directed at adult mosquitoes are insufficient for suppression of dengue and other Aedes-borne diseases [10 , 54] . A WHO dengue Scientific Working Group identified “analysis of the factors that contribute to the success or failures of national programs in the context of dengue surveillance and outbreak management” , including vector control , as a priority topic for future research [55] . Through two large-scale experimental studies and an assessment of a MoH emergency intervention campaign , our study evaluated an adulticiding strategy that is embedded in some national Aedes-transmitted virus control programs . We observed a clear Ae . aegypti population reduction during the extended period of repeated spray applications . These reductions were , however , not sustained after cessation of spraying . Our study design could not logistically include randomized replicates [30 , 51 , 56] because we focused on monitoring spraying in large neighborhoods of houses . A review of previous Ae . aegypti space spray studies [4] shows that each replicate included 50 or fewer houses so that movement of adults from surrounding houses could have impacted results . In contrast , we monitored spraying in large numbers of houses: more than 1 , 100 houses ( up to 2 , 100 houses ) during the two experimental interventions , and a MoH citywide emergency spray program . Our experimental design reduced the potential impact of adults moving into the sprayed sector from unsprayed locations . In the citywide spraying , all areas of the city were expected to have about the same decrease in Ae . aegypti densities so adult movement should not have impacted the recovery at all . There is clearly a tradeoff between degree of replication possible and the size of experimental units . In order to maintain study quality , our experimental interventions were supervised by trained entomologists . Our monitoring of the impacts of the L-2014 citywide emergency spraying provides a realistic and complimentary effectiveness assessment under practical , public health circumstances . It is also important to note that our study was primarily designed to provide data that could then be used to evaluate a computer simulation model [32] under extreme perturbation conditions , which was a major reason for evaluating a single centralized spray sector surrounded by a buffer sector . If viewed as an evaluation of a vector control measure , the design of our study has some important limitations: relatively short pre-intervention surveys , lack of randomization of intervention areas , and the absence of replicates in time . Cluster randomized trials ( CRTs ) are considered the gold standard for efficacy trials [30 , 57] . Yet properly powered studies are large , expensive , and logistically challenging , especially when measuring an impact on disease . When CRTs are not possible , large , carefully monitored studies such as ours can reassure public health agencies that employ emergency indoor space spraying programs . In addition , studies such as this one can provide target coverage goals and help manage expectations about the real-world impacts of spraying . Furthermore , studies of similar rigor are needed to evaluate the more commonly employed outdoor space spray campaigns . The effectiveness of pyrethroid applications varied between years , but was similar between citywide emergency sprays and experimental sprays in 2014 . Interestingly , in all experiments adult Ae . aegypti densities decreased significantly after the first cycle of spraying then fluctuated at relatively low levels during the remaining spray cycles: that is , additional cycles did not lower mosquito densities further . In all three interventions , adult populations partially recovered within 2 weeks of spray cessation . The pattern of rapid recovery of the Ae . aegypti population in our study is consistent with several previous reports [5] . Studies by Perich et al . [13 , 14] in Honduras and Costa Rica showed an approximately 90 percent reduction in adults one week after spraying , but the effect of the treatment was no longer significant after 6–7 weeks . In the two experimental suppression trials we could not definitively determine if recovery of population densities was from adults migrating in from the surrounding buffer sector and/or from new adults emerging from development sites within the spray sector . However , in the emergency citywide spraying , the recovery was similar to that in the experimental trials . This suggests that movement of adults was not the key factor . Mosquito densities after the L-2014 experimental spray were monitored for a longer period of time: 23-weeks post-spray in L-2014 versus 9-weeks post-spray in S-2013 . During L-2014 , the density of adults in the spray sector increased to well above that in the buffer . In L-2014 , ULV spraying resulted in a higher proportion of nulliparous females , indicating a shift to a younger adult female age distribution . This indicates that the spray sector continued to have active larval habitats that were producing new Ae . aegypti adults . In S-2013 , for example , 22 Ae . aegypti positive containers were identified in a single house during a post intervention survey , whereas the baseline survey of that house revealed only three containers total , of which only one was positive . This kind of variation illustrates the stochastic and dynamic nature of Ae . aegypti larval habitats [20 , 21] . The dramatic L-2014 post-treatment increase cannot , however , be explained by an outlier in the form of a “superproductive” household [33] . One possibility is compensation by the immature population due to a reduction in larval population densities , which led to reduced density dependent competition within containers and increased survival to adult emergence . This kind of rebound effect merits further investigation . In L-2014 , both emergency and experimental spraying had significant , but lesser impact on the adult densities than in S-2013 , even though L-2014 spray sector surveys with prior sprays were conducted ( on average ) fewer days after spray applications . The L-2014 24-hour mortality of caged sentinel mosquitoes was lower than in S-2013 , something that could be due to characteristics of the different insecticide used , changes in pyrethroid resistance levels in Iquitos mosquito populations between S-2013 and L-2014 , and/or differences in spray quality between the two experiments . By the end of 2014 , significant pyrethroid resistance was detected in Iquitos [39] . Although we did not detect pyrethroid resistance before the S-2013 experiment , we do not have similar assay information from populations evaluated just prior to the L-2014 experiment . It is possible , therefore , that the lower efficacy observed in the L-2014 experiment was due in part to resistance in the local Ae . aegypti population . By 2015 the MoH had abandoned use of pyrethroid insecticides for indoor spraying and switched to malathion in an effort to improve efficacy . A strong argument can be made that logistical challenges associated with application of ULV spray over a larger sector in the L-2014 experiment contributed to lower efficacy . First , Colt hand-held sprayers were only used in L-2014 when initially unsprayed houses were revisited , whereas in S-2013 they were used on at least 33% of the houses . Colt sprayers had significantly better and consistent droplet sizes than backpack sprayers . The L-2014 experiment was a much larger effort with at least double the number of backpack machines and MoH fumigators participating , and droplets were only evaluated on a fraction of the machines used . In addition , during the L-2014 experiment coverage rates were lower overall . Our results demonstrate that intensive , carefully administered space spraying can temporarily decrease the number and average age of female Ae . aegypti in houses . These results support smaller scale studies showing space spray induced reductions in Ae . aegypti density [12–15] . When , where , and how ULV mosquito control leads to meaningful reductions in disease remains a critical unanswered public health problem for policy makers . Computer simulation models have been employed to inform outcomes in limited situations , such as pathogen strain invasions ( e . g . Newton and Reiter [58] ) . Certain tentative recommendations , however , can be made based on existing data . Emergency indoor ULV spray interventions have the potential to mitigate Ae . aegypti-transmitted viruses , but coverage must be maximized with multiple spray cycles per house; i . e . , at least 3 spray cycles based on our experience in Iquitos [27] . Officials should have no expectations of sustained reductions in mosquito densities and must recognize that these sprays only have the potential to mitigate the immediate impact of an arbovirus outbreak . Quality control of spraying efforts and insecticide resistance testing must be an integrated component of national programs . Although these are not new messages [47 , 59] , our study adds new data to the vector control evidence base that we hope will better inform intervention programs and , thus , help refine policy for the application of space spray as a public health response to Ae . aegypti-transmitted viruses . | Aedes aegypti is a primary vector for medically important viruses that typically resides within houses . Indoor , ultra low volume ( ULV ) adulticide space spraying is considered to be more effective in controlling Ae . aegypti populations than outdoor spraying , and is widely used in tropical cities . Given the widespread use of indoor ULV spraying in emergencies by municipal control programs , the lack of large spatial scale evaluations is problematic . We conducted two large-scale experiments to evaluate indoor ULV pyrethroid spraying in the city of Iquitos , Peru in 2013 and 2014 , and we also evaluated a municipal spraying effort . Our results demonstrate that densities of adults can be reduced by ULV spraying , but that adult densities in sprayed areas return to approximately pre-spray densities in less than a month . These findings agree with results from previous , smaller scale experiments , and confirm that ULV spraying should be viewed as causing a short-term decrease of Ae . aegypti populations . We provide extensive detail regarding our experimental design and data collection so that our results can assist in establishing best practices for future assessments of ULV spraying efforts , as well as aid in testing predictions of mathematical models of Ae . aegypti population dynamics . | [
"Abstract",
"Introduction",
"Methods",
"and",
"materials",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"animals",
"research",
"design",
"surveys",
"population",
"biology",
"infectious",
"disease",
"control",
"insect",
"vectors",
"zoology",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
... | 2018 | Efficacy of Aedes aegypti control by indoor Ultra Low Volume (ULV) insecticide spraying in Iquitos, Peru |
The association of anaemia with intestinal schistosomiasis and hookworm infections are poorly explored in populations that are not limited to children or pregnant women . We sampled 1 , 832 individuals aged 5–90 years from 30 communities in Mayuge District , Uganda . Demographic , village , and parasitological data were collected . Infection risk factors were compared in ordinal logistic regressions . Anaemia and infection intensities were analyzed in multilevel models , and population attributable fractions were estimated . Household and village-level predictors of Schistosoma mansoni and hookworm were opposite in direction or significant for single infections . S . mansoni was found primarily in children , whereas hookworm was prevalent amongst the elderly . Anaemia was more prevalent in individuals with S . mansoni and increased by 2 . 86 fold ( p-value<0 . 001 ) with heavy S . mansoni infection intensity . Individuals with heavy hookworm were 1 . 65 times ( p-value = 0 . 008 ) more likely to have anaemia than uninfected participants . Amongst individuals with heavy S . mansoni infection intensity , 32 . 0% ( p-value<0 . 001 ) of anaemia could be attributed to S . mansoni . For people with heavy hookworm infections , 23 . 7% ( p-value = 0 . 002 ) of anaemia could be attributed to hookworm . A greater fraction of anaemia ( 24 . 9% , p-value = 0 . 002 ) was attributable to heavy hookworm infections in adults ( excluding pregnant women ) as opposed to heavy hookworm infections in school-aged children and pregnant women ( 20 . 2% , p-value = 0 . 001 ) . Community-based surveys captured anaemia in children and adults affected by S . mansoni and hookworm infections . For areas endemic with schistosomiasis or hookworm infections , WHO guidelines should include adults for treatment in helminth control programmes .
Anaemia remains an intractable public health problem in sub-Saharan Africa ( SSA ) . Children younger than five years and pregnant women are the focus of most epidemiological surveys , as prevalence is highest amongst these groups [1–11] . In SSA , over 64% of children younger than five years and 55% of pregnant women are estimated to have anaemia [2] . Designing treatment packages for anaemia is complex , owing largely to its multifactorial aetiology . These causes include genetic haemoglobinopathies , haemorrhage , bacteraemia , deficiencies of iron or other micronutrients including copper , folate , vitamins A and B12 , and parasitic infections of malaria , schistosomiasis , and hookworm [12 , 13] . If left untreated , anaemia can reduce work capacity , decrease immunogenicity , and impair cognitive or motor development [3–6 , 14] . In this paper , we focus on the relationship between anaemia and intestinal helminths . The relevance of intestinal schistosomiasis ( S . mansoni ) for anaemia remains controversial , including its causal mechanism [9 , 10] . Recent studies in SSA [15–17] focused on preschool children or adults in a single village and found no correlation between S . mansoni infection and anaemia . Yet , in school-aged children and pregnant women , heavy S . mansoni infection intensity has been shown to increase anaemia risk [18 , 19] . The relationship of hookworm infection and intestinal blood loss is well established [20] . This association is intensity-dependent with moderate to heavy infections required for anaemia [21–23] . In SSA , anaemia attributable to hookworm has been shown to be 18% in preschool children [21] , 5%-25% in school-aged individuals [22 , 23] , and 28% in pregnant women [24] . The associations of anaemia with S . mansoni and hookworm infections have been studied in the same population under two contexts: efficacy of anthelminthic treatment and polyparasitism . Amongst infected schoolchildren , separate treatments with praziquantel for S . mansoni and albendazole or mebendazole for hookworm have been found to reduce anaemia [19 , 25 , 26] . Although individually relevant in SSA , moderate to heavy co-intensities of S . mansoni and hookworm have been found to have no significant effect on anaemia in individuals under six years in Uganda [16] and school-aged children in Rwanda [27] . With previous epidemiological surveys in SSA focused on young children and pregnant women , current World Health Organization ( WHO ) guidelines for treating anaemia due to hookworm neglect community-based samples [11 , 28] . Additional community-based research is required to assess if guidelines should be established for treating anaemia attributable to S . mansoni infection [11 , 28] . There is a need to identify the levels of S . mansoni and hookworm infection intensities associated with anaemia in populations , which are not limited to young children , school-based samples , or pregnant women . Moreover , it is unknown whether S . mansoni and hookworm infections influence anaemia prevalence in the same age group for community-based samples . To prioritize treatment strategies for anaemia , there also is a need to accurately quantify the proportion of anaemia attributable to S . mansoni and hookworm infections in community-based samples . Herein , we present a large-scale , community-based investigation of anaemia and intestinal helminths in SSA .
Data were collected from August-September 2013 in 30 villages of Mayuge District , Uganda . Annual and en masse treatments with praziquantel for S . mansoni and albendazole for hookworm began in the study area in 2003 and were scaled-up to all sub-counties in the study district in 2004 by the national helminth control programme [29 , 30] . An area within five kilometers of Lake Victoria in Mayuge District was chosen , as previous nationwide surveys [31 , 32] showed both S . mansoni and hookworm infection prevalence exceeds 50% for schoolchildren aged 5–21 years . In these surveys , the prevalence of coinfection was unknown . Accordingly , this area , which is within five kilometres of Lake Victoria in Mayuge District , was classified as a high-risk area for schistosomiasis infections . Every person aged five years and older was treated annually with praziquantel . If the area was not high-risk , praziquantel would only be available for children through primary schools . Adults are generally not treated through mass drug administration ( MDA ) with albendazole for hookworm infections [33] . However , our study area was endemic for lymphatic filariasis and all individuals aged five years and older were treated annually with a package of albendazole and ivermectin . MDA is the only substantial strategy for controlling these helminth infections , thus this context must be studied to reflect the true public health situation . The Ugandan national control programme aims to administer annual treatment for schistosomiasis and twice yearly treatment for soil-transmitted helminths . However , due to in-country delays , Mayuge District did not receive MDA for 1 . 5 years prior to our study . Thirty villages were selected for this study as follows . In February 2013 , 41 villages were visited in this catchment . For each of the eight sub-counties of Mayuge District , the district health officer identified one village with known S . mansoni transmission . At least four villages in closest physical proximity to these initial eight villages were sampled . One cluster of 3–8 villages was sampled in each sub-county . In addition to the collection of village waypoints , researchers drew maps with village chairmen to identify locations of taps , rice paddies , standing seasonal water , public latrines , homes , and primary schools . These maps were used to select six village clusters from the initial eight sub-counties that were similar in size , language , infrastructural development , quality of homes , and distance to established towns . The sample included 1 , 832 individuals aged 5–90 years ( Mean 24 . 31 , Std . Dev . 16 . 90 , includes one four year-old ) from 916 households . Village health team members selected at least 30 households and two people per household in their village . To ensure men participated , health team members were instructed to stratify participants by gender and familial position , e . g . mothers with daughters , mothers with sons , fathers with daughters , and fathers with sons . This stratification enabled a multilevel analysis ( described below ) to account for latent , unobserved household or village effects on anaemia . Child participants were required to be at least five years , which is the minimum age to treat S . mansoni infections with praziquantel [33] . Moreover , sampling of children focused on ages 10–14 , which is where the peak intensity of S . mansoni infection occurs in most endemic areas [34] . There were no other restrictions on the child , who was chosen by the village health team member . Lastly , at-risk groups of fishermen ( S . mansoni ) and farmers ( hookworm ) were sampled . Stratification by occupation was conducted to ensure representation of occupations that were characteristic of the high prevalence villages . This approach was needed , as fishermen and farmers leave the village during the day to work and may otherwise be underrepresented . This sampling did not affect the internal validity of the data and should not have an effect on the generalizability of the study if endemic , high-risk villages are of interest . However , given the number of villages , a complete demographic census was not performed and should be considered in future work . A random non-stratified sampling strategy may be preferred for areas with low S . mansoni or hookworm infection prevalence . Each participant provided one stool sample . Stool samples were immediately processed upon receipt and standard Kato-Katz methodology for two thick smear slides ( 41 . 7 mg ) [35] was used . Technicians read slides within 30 minutes of preparation for the presence of hookworm eggs and slides were reread 24 hours after preparation to count eggs of S . mansoni and other STHs . Ten percent of slides were retained for S . mansoni quality control egg counts by a senior technician , who confirmed the accuracy of the field readings . Egg counts were multiplied by 24 then averaged to represent S . mansoni and hookworm infection intensities as the number of eggs per gram ( EPG ) . To assess anaemia , every participant provided a fingerprick blood sample for the measurement of haemoglobin ( Hb ) ( g/L ) in Hemocue Haemoglobinometers ( Radiometer Group , Sweden ) . A short questionnaire was administered to all adults , who also answered on behalf of the child . Participants provided demographic information , the last incident of malaria , and income-earning occupation ( if any ) of the household head . The data were analyzed in Stata v . 12 . 1 . The main variables used to examine anaemia were as follows . WHO categories for light ( 1–99 EPG ) , moderate ( 100–399 EPG ) , and heavy ( 400+ EPG ) S . mansoni infection intensity were used [36] . These values also were applied to classify hookworm , as 400 or more EPG represented the top 10th percentile of hookworm infection intensity measured in our study . This method of hookworm infection classification also is employed in Pullan et al . [22] . Anaemic individuals were classified according to WHO guidelines for Hb by age and gender [11] . All models in this paper excluded basic age-gender-infection or S . mansoni-hookworm interactions , as there was insufficient support of model improvement from likelihood ratio tests ( p-value>0 . 05 ) . Descriptive statistics of variables used in analyses are provided in Tables A-D in S1 File . To understand the distribution of infection intensities and to check the robustness of this data with known epidemiological studies , the risk factors of S . mansoni and hookworm infections were compared in ordinal logistic regressions [37] . Infection categories were ranked from uninfected to heavy infections and used as dependent variables . In these models , the relevance of age , gender , household head occupation , and village-level factors were examined for infection intensity . Village predictors included a continuous variable for the number of houses and the following binary indicators for environmental factors within the village: beach on Lake Victoria , lake site only ( no beach , only small boat landing site to Lake Victoria ) , a rice paddy ( rice farm ) , and three or more village roads . Additionally , a binary indicator was used that was equal to one if the distance of the village centre was greater than 0 . 50 kilometres to Lake Victoria . Standard errors were adjusted with robust sandwich estimators at the household level [38] . Ordinal logistic regressions present cumulative probabilities and assume proportional odds of being in higher outcome categories . There was sufficient evidence that age , the occupations of business and rice farmer , and the distance of the village centre to Lake Victoria violated this assumption in the hookworm model [39] . Using methods described in Peterson and Harrell [40] and Williams [39] , these variables were allowed to vary across hookworm infection levels and the associated beta estimates were provided . Anaemia status was analyzed with a multilevel logistic regression that controlled for household and village variance . General linear and latent mixed models were used [41] . Infection categories were incorporated as predictors of anaemia and presented as binary variables of low , moderate , and heavy S . mansoni or hookworm infections ( with uninfected as the base category ) . Gender , dummies for household head occupation , and age ( in years ) were included as covariates . Other covariates of age grand-mean-centered squared , self-reported malaria ( individual received antimalarial medicines from government or private health clinic within the past six months ) , and village-level factors were tested by forward selection . Univariate and empty logistic models were compared . If the likelihood ratio test was significant ( p-value<0 . 05 ) , the variable was used ( Table E in S1 File ) . A crude global R2 was calculated using the square of the correlation between fitted and actual anaemia values [42] . Although this method is used in linear regression , the crude global R2 does not provide insight into the proportional reduction in error variance explained by the full model . Accordingly , more robust calculations of R2 were calculated using recent methods of Nakagawa and Schielzeth [43] . For variation explained by only the fixed component and full model , the marginal R2 and conditional R2 , respectively , were estimated . Intra-class correlation ( ICC ) coefficients were calculated as described in McGraw and Wong [44] . The assessment of coinfection and anaemia only differed from the aforementioned anaemia status model with respect to the infection variables used; otherwise , all other variables and model specifications were the same . S . mansoni and hookworm coinfection categories were presented as binary indicators and used as predictors of anemia . The base category for each coinfection indicator included no infection ( zero EPG ) and single infections of S . mansoni or hookworm . Four categories were used to classify coinfection intensity . Light , moderate , and heavy coinfections were defined , respectively , as at least 1–99 EPG , 100–399 EPG , and 400+ EPG of both S . mansoni and hookworm infections . To quantify the impact of S . mansoni and hookworm infection on anaemia , population attributable fractions ( PAF ) were estimated using maximum likelihood procedures described in Greenland and Drescher [45 , 46] . This method adjusts for covariates and calculates PAF confidence intervals ( CI ) , but random effects or latent variable distributions in the data cannot be preserved . Accordingly , logistic regressions were used with the same variables as the multilevel anaemia model and clustered standard errors at the household level [38] . Infection intensities found significant in the anaemia logistic model were examined . Two PAF scenarios were estimated: elimination and intensity reduction . These scenarios were assessed across the whole study sample , a sub-population of heavily infected individuals , and a sub-group that excluded school-aged children and pregnant women . This study was reviewed and approved by the Uganda National Council of Science and Technology ( SS3082 ) , Office of the President in Uganda ( SS3082 ) , and Cambridge University Human Biological Research Ethics Committee ( HBREC2013 . 10 ) . Written informed consent was obtained from participants or their guardians . For adults or guardians who indicated they were unable to write or who preferred to provide fingerprints , verbal informed consent and a fingerprint signature were obtained ( reviewed and approved in SS3082 and HBREC2013 . 10 ) .
Prevalence ( at least one detectable EPG ) of S . mansoni and hookworm was 36 . 4% ( 667/1832 ) and 40 . 5% ( 741/1832 ) , respectively , which included 13 . 1% ( 240/1832 ) of the population with coinfection . Overall anaemia prevalence was 44 . 4% ( 813/1832 ) . Anaemia prevalence coincided with the highest average S . mansoni or hookworm EPG ( Tables C and D in S1 File ) . As shown ( Fig 1 ) , anaemia prevalence formed a U-curve , being highest in children and the elderly . These age groups harboured the heaviest S . mansoni and hookworm infections , respectively . Anaemia prevalence was at least 34 . 6% in all age groups . Although coinfections of S . mansoni and hookworm were observed in 13 . 1% of the population , only 0 . 55% ( 10/1832 ) of the population had coinfections with heavy infection intensity of both S . mansoni and hookworm ( Table 1 ) . At the village level , mean S . mansoni EPG and mean hookworm EPG were inversely related ( Spearman rho -0 . 407 , p-value<0 . 0001 ) . Similarly , most communities did not have high prevalence of both infections ( Spearman rho -0 . 562 , p-value<0 . 0001 ) . In ordinal logistic models in Table 2 , S . mansoni and hookworm showed no overlap of infection determinants . Any variables that were significant in both models were opposite in direction . A one-year increase in age or being female decreased the likelihood of S . mansoni infection by 1 . 98% ( p-value<0 . 001 ) and 31 . 5% ( p-value<0 . 001 ) , but increased the odds of hookworm infection by 1 . 2% ( p-value<0 . 001 ) and 42 . 7% ( p-value<0 . 001 ) , respectively . Belonging to a community where the village centre is farther than 0 . 50 kilometres from Lake Victoria decreased the likelihood of S . mansoni infection by 67 . 9% ( p-value<0 . 001 ) and increased the probability of hookworm infection by 48 . 9% ( p-value = 0 . 001 ) . For S . mansoni , variables measuring water contact increased the probability of infection . Fishing occupations increased the likelihood of S . mansoni 2 . 70 fold ( p-value<0 . 001 ) when compared to households with an unemployed person or subsistence farmer as the household head . Individuals living in a village with a beach were 2 . 06 times ( p-value<0 . 001 ) more likely to have S . mansoni infections than individuals in a village without a freshwater site . Additionally , having more homes in the village increased the odds of heavy S . mansoni infection by 57 . 2% ( p-value<0 . 001 ) . Predictors that were only significant for hookworm infection included , at the household level , business ownership and , at the village level , the number of roads . Belonging to a household where the household head owned a business reduced the odds of heavy hookworm infection by 65 . 4% ( p-value<0 . 001 ) when compared to subsistence farmers or the unemployed . Having more than two roads , which was a proxy indicator for the spatial spread or low density of households in a village , increased the likelihood of hookworm infections by 93 . 7% ( p-value<0 . 001 ) . Table 3 presents the association of S . mansoni and hookworm infection intensity with anaemia . The variance components model is provided in Table F ( S1 File ) . Individuals with moderate S . mansoni infection intensity were 56% more likely ( p-value = 0 . 034 ) to have anaemia than uninfected individuals . Heavy S . mansoni infections increased anaemia risk 2 . 861 fold ( p-value<0 . 001 ) when compared to participants with no detectable S . mansoni infection . The probability of anaemia was 65% higher ( p-value = 0 . 008 ) amongst individuals with heavy hookworm infection intensity when compared with uninfected individuals . These findings were significant despite older age ( OR 0 . 98 , p-value<0 . 001 ) decreasing the likelihood of anaemia . Only the effect of moderate S . mansoni infection intensity on anaemia was lost when analyzed against Hb ( Table G in S1 File ) . Table 4 presents the association of anaemia with S . mansoni and hookworm coinfection . Individuals with light , moderate , or heavy coinfection intensity were not significantly at risk of anaemia ( p-value>0 . 05 ) when compared to participants with no infection or single infections . These results remained robust when coinfection intensities were analyzed against Hb concentration ( Table H in S1 File ) . As a robustness check , a binary indicator was examined for ‘any coinfection’ , which was defined as a participant with at least one EPG of S . mansoni and hookworm . This indicator also was insignificantly related to anaemia and Hb ( p-value>0 . 05 ) when compared to participants with no infection or single infections ( Table I in S1 File ) . Table 5 presents the PAF for S . mansoni and hookworm infection intensities that were found significant for anaemia in the logistic model of Table J ( S1 File ) . Without household and village variation controlled , moderate S . mansoni was borderline insignificant ( p-value = 0 . 056 ) against anaemia . In scenario one ( Panel 1A ) , for all individuals , elimination of heavy S . mansoni infection could reduce anaemia by 3 . 6% ( 95% CI 2 . 1% , 5 . 0% ) , whereas elimination of heavy hookworm infections could decrease anaemia by 3 . 0% ( 95% CI 1 . 0% , 5 . 0% ) . For adults ( excluding pregnant women ) , only 2 . 4% ( 95% CI 1 . 4% , 3 . 4% ) of anaemia was attributable to S . mansoni infections . Yet , over 4 . 5% ( 95% CI 2 . 7% , 6 . 3% ) of anaemia for school-aged children , who had the highest infection intensity , and for pregnant women was attributable to S . mansoni infections . The proportion of anaemia attributable to hookworm infections was greater in age groups with the highest hookworm infection intensities . Adults had a greater fraction of anaemia attributable to hookworm ( PAF 4 . 6%; 95% CI 1 . 4% , 7 . 6% ) when compared to school-aged children and pregnant women ( PAF 1 . 8%; 95% CI 0 . 6% , 3% ) . In Panel 2 , two subsets of the population with either heavy S . mansoni or heavy hookworm infections were used to calculate the population attributable fractions amongst the infected ( PAFI ) . The elimination of S . mansoni in individuals with heavy S . mansoni infection could reduce anaemia by 32 . 0% ( 95% CI 21 . 5% , 41 . 1% ) . Hookworm elimination amongst people with heavy hookworm infection could decrease anaemia by 23 . 7% ( 95% CI 9 . 4% , 34 . 7% ) . The effects of eliminating infections were further examined by age for individuals with heavy infection intensities . Heavily infected adults , excluding pregnant women , had a larger fraction of anaemia attributable to both helminth infections when compared to heavily infected school-aged children and pregnant women . Curing heavily infected , non-pregnant adults could decrease 37% ( 95% CI 25 . 2% , 47% ) and 24 . 9% ( 95% CI 10% , 37 . 4% ) of anaemia attributable , respectively , to heavy S . mansoni and hookworm infections . Eliminating infections in heavily infected school-aged children and pregnant women may reduce 30 . 4% ( 95% CI 20 . 3% , 39 . 2% ) and 20 . 2% ( 95% CI 8 . 5% , 30 . 4% ) of anaemia due to heavy S . mansoni and hookworm infections , respectively . In scenario two , the PAF was calculated using a strategy of only reducing infection intensity as opposed to curing individuals . Covariates were accounted for and a model was used similar to that used in Table J ( S1 File ) except indicators were excluded for low and moderate S . mansoni and hookworm infections ( Table K in S1 File ) . This analysis presented the potential impact of reducing heavy S . mansoni infections to uninfected/low/moderate S . mansoni infections and reducing heavy hookworm infections to low/uninfected/moderate hookworm infections . Decreasing intensity would have an impact on anaemia for the full sample , school-aged children and pregnant women , or only adults ( excluding pregnant women ) that is comparable to the elimination of these infections from these groups . PAFs and PAFIs only differed , respectively , by 0 . 01–0 . 3% and 0 . 8–1 . 9% ( Panels 1B , 2B in Table 5 ) .
With an increase in the number of anaemia cases from the year 1990 to 2010 in SSA [1] , a better understanding is needed of how to simplify monitoring and treatment of the multiple causes of anaemia . In this paper , we assessed the associations of S . mansoni and hookworm infections with anaemia . A community-based sample was used of 1 , 832 individuals aged 5–90 years across 30 villages in Uganda . When examined against age , peak infection intensities coincided with the highest prevalence of anaemia . The heaviest intensity of S . mansoni was in children and the highest intensity of hookworm was in the elderly . Predictors of infection risk including age , gender , occupation , and village-level factors were significant for only one infection or opposite in direction for S . mansoni and hookworm . Differences in S . mansoni and hookworm risk factors have been shown in studies of single infections [47 , 48] . To our knowledge this study is the first to show that S . mansoni and hookworm infections influence anaemia in separate groups of individuals within one study setting . In community-based samples , guidelines to target individuals requiring treatment for anaemia due to S . mansoni or hookworm infections must delineate separate profiles for each infection . Yet , community-based sampling can be used to capture diverse causes of anaemia due to both intestinal helminths . Our study was able to identify the relative contribution of S . mansoni and hookworm infections to anaemia despite potential confounders of malnutrition and malaria . Malnutrition is an important aspect of anaemia risk [12] . For malnutrition to be a confounder in our analysis , it must have an individual-level effect on anaemia risk . The multilevel model in this study accounted for latent variation in anaemia attributable to household and village variance . Accordingly , important unobserved differences between households or communities for malnutrition were considered in these latent factors . For example , household factors that affect anaemia may include maternal care for children , household diet and meals , or socioeconomic factors whereas community factors may include sanitation infrastructure or physical proximity to a health centre [49 , 50] . At the individual level , Pullan et al . finds significant associations of S . mansoni and hookworm infections with anaemia after controlling for malnutrition in schoolchildren in Kenya [22] . Self-reported malaria , which indicated if study participants received treatment for malaria in the six months preceding the study , was uncorrelated with anaemia . For malaria to be a confounder in our study , two issues must be observed . Formal diagnoses of malaria must conflict with self-reported treatment and a specific set of individuals must be coinfected , i . e . individuals with heavy S . mansoni or heavy hookworm infection intensities . Coinfections of P . falciparum and S . mansoni or hookworm have been shown in SSA to increase anaemia prevalence in children when compared to single S . mansoni or hookworm infections [51 , 52] . Such coinfections have been shown to have a prevalence of 26 . 47% in children aged 10–14 years in our study area and are a limitation of this study [53] . Another factor to consider is if malaria prevalence decreases with the distance to freshwater bodies , as was the case with S . mansoni infections . However , this confounder is unlikely in our study setting since individuals within the same village were only 2 . 1% more likely to have anaemia when compared to individuals of other villages ( ICC , Table 3 ) . Also , the presence of swamps in a village and the distance of the village centre to Lake Victoria was investigated and found insignificant ( p-value>0 . 05 ) for anaemia risk ( Table 3 , Table E in S1 File ) . Concerning other coinfections , additional intestinal helminths were directly examined as potential predictors of anaemia . The infection prevalence of trichuriasis was 1 . 04% and ascariasis was 0 . 11% in our study villages , but no associations with anaemia were found ( Spearman correlations: Trichuriasis rho 0 . 028 , p-value = 0 . 234; Ascariasis rho -0 . 030 , p-value 0 . 207 ) . The relationships of S . mansoni and hookworm infections with anaemia were intensity-dependent; heavy infection intensities ( 400+ EPG ) significantly increased anaemia risk . Individuals with heavy S . mansoni were 2 . 86 times more likely to have anaemia than uninfected participants . This effect amongst our community-based sample is considerably larger than the influence of heavy S . mansoni infection in individuals sampled from Ugandan primary schools , who had increased anaemia risk of 1 . 57 fold [19] . Further work is needed to identify if community-based samples best include children at the highest risk of anaemia from S . mansoni infection . Community-based samples , such as our study , include many non-enrolled children and heavy infections may be concentrated in this group . Although beyond the scope of this study , there is a need to identify the major mechanism of S . mansoni infection that causes anaemia . S . mansoni may be in direct competition with the human host for iron in the intestinal and blood stages of the parasite [54] . Alternatively , blood loss could occur from intestinal bleeding when eggs need to traverse the epithelial , mucosal barrier [10 , 15 , 55] . If the intestinal epithelium is compromised then host intestinal iron absorption can be impaired [55] . Additionally , if severe S . mansoni infection elevates markers of inflammation such as Interleukin 6 , then a feedback mechanism could occur that inhibits host intestinal iron absorption through up-regulation of hepcidin [56] . In addition to Hb counts , to address these possible causes , future studies must measure occult blood and indicators of iron metabolism such as serum ferritin and soluble transferrin receptors [15 , 56] . Heavy hookworm infections increased anaemia risk 1 . 65 fold when compared to people with no detectable hookworm infection . This estimate concurs with existing knowledge of moderate and heavy hookworm infection intensity and anaemia [7 , 8] . The effect of hookworm infection may appear less than the influence of S . mansoni infection on anaemia . However , in this study , heavy hookworm infection intensity was classified as the top 10th percentile of hookworm infections ( 400+ EPG ) . Although 400 or more EPG is the WHO classification of heavy S . mansoni infection , this intensity is considered by the WHO as a light hookworm infection [33] . Pullen et al . [22] shows very light hookworm infections in Kenyan schoolchildren ( 100 EPG ) are associated with anaemia . In this paper , by assessing entire communities , a sensitive detection was possible of the relative contribution of hookworm infection to anaemia for every age group . We showed that the elderly are vulnerable to anaemia with only light hookworm infections . With morbidity detectable from only light hookworm infections , future research is warranted to identify the hookworm species present in the study area . Hookworm virulence can vary depending on the species . Albonico et al . has shown that Ancylostoma duodenale , although less prevalent than Necator americanus , can cause greater iron-deficiency anaemia [57] . S . mansoni and hookworm coinfections were insignificant for anaemia when compared to no infection or single infections . This result is consistent with studies in Uganda and Rwanda that assessed S . mansoni and hookworm coinfections in children [16 , 27] . Coinfection was insignificant in our study because only heavy infection intensity was associated with anaemia , and heavy S . mansoni and heavy hookworm infections did not overlap by age , socioeconomic , or environmental characteristics . If repeated in other settings , this finding can provide explanations for why other studies in SSA [16 , 27] also have failed to find associations between anaemia and coinfections of S . mansoni and hookworm . Treatment of anaemia must be made a priority for individuals with heavy S . mansoni and heavy hookworm infections . A large proportion of anaemia amongst infected individuals was attributable to intestinal helminths . S . mansoni explained 32 . 0% of anaemia in individuals with heavy S . mansoni infection intensity . For people with heavy hookworm infections , 23 . 7% of anaemia could be attributed to hookworm infections . When the anaemia attributable to helminth infections was further analyzed by age , the proportion of anaemia due to S . mansoni or hookworm infections was greater in sub-groups with higher average infection intensities . Children had a greater fraction of anaemia attributable to S . mansoni infections than adults and non-pregnant women . However , adults ( excluding pregnant women ) had a greater proportion of anaemia due to hookworm infections when compared to school-aged children and pregnant women . These findings concur with our results that show anaemia prevalence peaks in age groups that have the highest infection intensities . What is striking about our results is that when only heavily infected individuals were examined by age , the anaemia attributable to both S . mansoni and hookworm infections was highest amongst adults . This finding raises an important question for future studies . Do children have more competing , multifactorial causes of anaemia than adults ? If so , how does competition amongst causes of anaemia affect the morbidity caused by helminth infections ? Most importantly , amongst heavily infected individuals , adjusted PAF analysis suggested that reducing infection intensity would have an impact on anaemia of only 1–2% lower than efforts to eliminate heavy S . mansoni or hookworm infections . Hence , anthelminthic treatment , which is the only strategy widely implemented for controlling S . mansoni and hookworm infection intensities , is likely the most effective method to decrease anaemia attributable to these infections [19 , 25 , 26] . Preventive chemotherapies are widely distributed through MDA to treat human helminthiases and have been administered annually since 2003 in our study area [29] . Yet , within the context of MDA , we found that widespread morbidity persists due to intestinal helminths . With increased funding commitments and drug donations [58] as well as programmes progressing from infection control to elimination [59] , there is an imminent need to expand treatment coverage with PCs for communitywide anaemia . Schoolchildren , who bear the greatest burden of anaemia [2] , must remain a priority for the WHO . In addition , for areas endemic with schistosomiasis and hookworm infections , our study directly appeals to the WHO to establish guidelines to include adults for treatment in helminth control programmes [33] . Adult eligibility for treatment is limited . Praziquantel is only available for adults in high-risk communities , which include areas with greater than 50% schistosomiasis infection prevalence in schoolchildren . For hookworm infections , adults mainly receive indirect treatment . When communities are targeted for lymphatic filariasis , all individuals aged five years and older are eligible for a package of albendazole and ivermectin . Only women of childbearing age and a few occupations including tea pickers and miners are eligible for hookworm treatment [33] . Our study suggests that guidelines for treating hookworm infection should not be limited to children and pregnant or potentially pregnant women . We demonstrated that adults exhibit serious morbidity associated with helminth infections . The availability of treatment for the elderly is of particular concern . By harbouring the heaviest hookworm infections , the elderly are a main reservoir of reinfection for not only schoolchildren , but also entire communities . If community-based epidemiological surveys are repeated in other countries , our findings can help consolidate WHO guidelines for use of preventive chemotherapies in SSA to treat anaemia caused by intestinal helminths . | Anaemia remains an intractable public health problem in sub-Saharan Africa ( SSA ) , owing largely to its complex and multifactorial causes . The absolute number of anaemia cases has increased by 50% in SSA from the year 1990 to 2010 . To administer the appropriate treatment , identification of anaemia attributable to specific causes is required . In this paper , we assessed the relative contribution to anaemia of intestinal schistosomiasis ( Schistosoma mansoni ) and hookworm ( a soil-transmitted helminth ) . The effect of S . mansoni on anaemia is controversial; this infection has not been found to consistently affect anaemia . However , hookworm has widely been shown to increase anaemia risk . Yet , limited information is available about the effect of S . mansoni and hookworm infections in populations that are not limited to children or pregnant women . Without this information , current World Health Organization guidelines for the treatment of anaemia due to intestinal helminths neglect a large portion of the general population by age and gender . To address this lack of information , we conducted a communitywide assessment of the effects of S . mansoni and hookworm infections on anaemia in rural Uganda . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Influence of Schistosoma mansoni and Hookworm Infection Intensities on Anaemia in Ugandan Villages |
Forward genetics screens with N-ethyl-N-nitrosourea ( ENU ) provide a powerful way to illuminate gene function and generate mouse models of human disease; however , the identification of causative mutations remains a limiting step . Current strategies depend on conventional mapping , so the propagation of affected mice requires non-lethal screens; accurate tracking of phenotypes through pedigrees is complex and uncertain; out-crossing can introduce unexpected modifiers; and Sanger sequencing of candidate genes is inefficient . Here we show how these problems can be efficiently overcome using whole-genome sequencing ( WGS ) to detect the ENU mutations and then identify regions that are identical by descent ( IBD ) in multiple affected mice . In this strategy , we use a modification of the Lander-Green algorithm to isolate causative recessive and dominant mutations , even at low coverage , on a pure strain background . Analysis of the IBD regions also allows us to calculate the ENU mutation rate ( 1 . 54 mutations per Mb ) and to model future strategies for genetic screens in mice . The introduction of this approach will accelerate the discovery of causal variants , permit broader and more informative lethal screens to be used , reduce animal costs , and herald a new era for ENU mutagenesis .
Forward genetic screens in mice carrying mutations introduced by the alkylating agent ENU can provide important and entirely novel insights into gene function [1] , [2] , [3] , [4] . This approach does not require any prior assumption about mechanism , and by inducing random point mutations ENU generates viable phenotypes that mimic human disease . In the classic approach mice treated with ENU are bred to generate pedigrees segregating thousands of mutations , which are screened for phenotypes of interest . However determining which of the many induced mutations underlies the phenotype is a significant bottleneck in the process , requiring additional generations of breeding and outcrossing to another inbred laboratory strain in order to generate a linkage map , followed by sequencing of candidate genes or regions . This process is time consuming and costly . For example conventional fine mapping to obtain a linkage region of around 3 Mb ( 20–30 genes ) requires at least 2 generations of additional breeding and genotyping 100–200 markers in 30–60 F2 mice . The need to propagate the mice tends to require non-lethal screens , which limits the range of assays and the scope to detect phenotypes . Furthermore , outcrossing can introduce unseen confounding variants affecting the trait , and tracking the phenotype through additional generations is complicated and can be unreliable [5] . Although whole genome and exome sequencing offer the prospect of accelerating discovery , current strategies remain dependent on conventional mapping [6] , [7] , [8] , [9] . In this study we address the ENU bottleneck by showing how it is possible to use WGS and identity by descent to isolate a causative mutation rapidly and efficiently on a pure , single strain background , without the need for outcrossing or additional breeding . The strategy will allow the use of lethal and more informative screens . It will accelerate the discovery of new variants , permit a greater focus on novel mutations , and make forward genetics more accessible by reducing costs and broadening the screens .
In a typical strategy for generating and screening ENU mutant mice , C57BL/6J ( B6 ) ENU-treated founders are bred with B6 females to generate G1 founders , establishing pedigrees in which pairs of G2 mice produce G3 mice segregating recessive and dominant mutations ( Figure 1A ) . To assess the utility of WGS in the analysis of such mice , we chose a pedigree identified as ENU16CH17a where the recessive phenotype was peripheral B cell lymphopenia ( Figure 1B and Figure S1 ) , and performed WGS on three affected G3 mice from a single G2 pair to high coverage ( average 24× per individual ) . The causative variant in ENU16CH17a will belong to a haplotype that is shared by all the sequenced mice and inherited from an ENU-treated ancestor . By constructing chromosomal maps of the homozygous variants in all three animals we could demonstrate clustering of mutations within haplotype blocks inherited from ENU-treated founders ( Figure 1C and Figure 1D ) . This suggested that the identification of IBD regions in multiple mice would be an elegant and efficient approach for mapping and identifying causative mutations . We developed a method based on the Lander-Green algorithm which uses genetic markers , knowledge of the pedigree and recombination rates , to infer the flow of alleles through the genealogy [10] . We modified the algorithm to exploit the partial knowledge of the G1 founder genotypes . Our implementation uses probabilistic variant calls to identify haplotypes from the four founder mice ( ENU1 , ENU2 , WT1 and WT2 ) across the genome of each G3 individual ( Figure 2A and Materials and Methods ) . The rate of ENU induced mutation is low and must be distinguished from both WGS artifacts and background variation from the reference , so we developed a series of filters to exclude non-ENU variants ( Materials and Methods and Figure S2 ) . The haplotypes assigned by our algorithm identify the IBD regions ( Figure 2B ) . IBD homozygous regions comprise 95 . 3 Mb ( 3 . 6% of the genome ) , containing 137 variants , including only two mutations in coding regions , both on chromosome 4 ( Figure 1D ) . One mutation at position 3 , 710 , 143 was an A to G mutation inducing a missense change in the Src-kinase encoding gene Lyn ( Figure 1E ) . The mutation corresponds to a threonine to an alanine substitution at amino acid residue 410 in exon 12 within the highly conserved Src activation loop in the protein kinase domain ( Figure 1F ) . The phenotype seen in ENU16CH17a has been described in mice carrying a threonine to lysine substitution at the same codon in Lyn [11] , indicating that the mutation is causative . The other mutation , at position 66 , 590 , 107 , encoded a missense mutation in a single transcript of Toll-like receptor 4 reported by Ensembl ( Tlr4-004 ENSMUST00000107365 ) , but absent from the RefSeq dataset . Tlr4 deficient mice do not have obvious defects in B cell development [12] . The ENU16CH17a pedigree carries a recessive functional mutation; however , to demonstrate the wider application of our method for dominant traits , we examined shared IBD heterozygous ENU mutations in the same G3 mice . The 3 ENU16CH17a mice share one or more haplotypes from a common ENU founder across regions comprising 40 . 8% of the genome ( 1083 . 8 Mb ) ( Figure 2B ) , containing 26 heterozygous candidate mutations shared by all 3 mice , comprising 25 missense and one splicing mutation; there were no nonsense mutations . PolyPhen-2 [13] predicted 9 as benign , leaving 17 heterozygous shared mutations with possibly deleterious effects . Sanger sequencing confirmed the presence of all 28 homozygous and heterozygous IBD variants ( Table S1 and Table S2 ) . The frequency of ENU mutations is dose related [14] , [15] and may differ according to the mouse strain [16] . However , previous estimates of the ENU mutation frequency , which have ranged from 0 . 5 to 10 , have been confounded by small datasets and locus specific bias [17] , [18] , [19] , [20] , [21] , [22] , [23] . By observing the frequency of variants in the homozygous ENU regions and subtracting the background rate observed in homozygous WT regions , we have calculated the ENU mutation rate in the ENU16CH17a pedigree to be 1 . 54 mutations . We excluded errors due to assignment of homozygous regions or inadequate coverage by modeling the effect of expansion or contraction of regions and of reduction in coverage ( Figure 3A and 3B ) . The estimate of mutation frequency was also insensitive to changes in the assumed mutation frequency used in the algorithm to predict IBD regions - with assumed mutation frequencies in the range 0 . 25 to 3 . 0 mutations , the estimated ENU frequency remained between 1 . 52 and 1 . 58 mutations ( Figure S3A ) . Within the homozygous ENU regions , we could confirm the well-described transition∶transversion ratio and A-T base preference of ENU induced mutations [24] , [25] . We found a 1 . 50∶1 transition: transversion ratio in ENU mutations compared to a 2 . 17∶1 ratio in naturally occurring mouse SNPs ( Figure 3C ) . We confirmed the distinctive base preference signature of ENU mutations , which is mainly due to error-prone repair of and ethylthymidine leading to AT to TA transversions ( 28 . 5% of our mutations ) , and AT to GC transitions ( 45 . 0% of our mutations ) , respectively [26] , [27] ( Figure 3D ) . We found 78 . 7% of all homozygous mutations were at A:T sites , compared to the 58% AT content of the mouse genome [28] and different to non-ENU variants seen in homozygous WT regions , of which 39 . 5% were at AT sites ( Figure 3E ) . For a high throughput ENU program , an analysis based on IBD would ideally identify SNPs and IBD regions accurately even at low coverage per individual . Therefore to model this we simulated a lower coverage dataset by randomly selecting subsets of ENU16CH17a reads and checked the consistency of variant calls at different simulated levels of coverage compared to 24× per mouse . The assignment of IBD regions remained highly consistent with the complete dataset down to very low coverage levels ( Figure 4A and Figure S4 ) . At 5× coverage per mouse , 93% of homozygous and 91% of heterozygous IBD regions seen at 24× were assigned , and 83% of homozygous and 77% of heterozygous variants in IBD regions overlapped with those found in IBD regions at full coverage ( Figure 4B ) . Within the validated set of non-synonymous coding and splice site mutations , we identified all the homozygous IBD variants ( 2/2 ) , and 69% of heterozygous IBD variants ( 18/26 ) at 5× ( Figure 4B ) . We compared our IBD approach to a simple method of selecting all variants observed in all 3 affected mice . At lower coverage levels the simple approach identified very large numbers of homozygous shared variants compared to the IBD method; at 5× coverage there were 586 shared variants compared to 158 by IBD ( Figure 4C ) . This error is likely to be due to the miscalling of heterozygous variants as homozygous , coupled with accumulation of further heterozygous variants , since the overall number of heterozygous shared variants at each simulated depth is greater than that observed using IBD ( Figure 4D ) . By only considering variants in IBD genomic intervals , we distinguished homozygous from heterozygous variation more accurately and reduced the number of spurious shared variants , making isolation of causative mutations feasible at low coverage . To confirm the utility of our method at low coverage experimentally , we Sanger sequenced shared variants from 3 affected mice in a second ENU pedigree , which had been sequenced by WGS at 4× per mouse . We again identified the causative recessive mutation , and found true-positive rates of 85 . 7% ( 24/28 ) for homozygous variants and 86% ( 48/56 ) for heterozygous variants . Within the subset of coding variants , we identified 100% ( 4/4 ) of the homozygous variants , including the causative variant , and 96% ( 24/25 ) of the heterozygous variants ( Table S4 ) . As expected , Sanger sequencing of variants from non-IBD regions revealed lower true-positive rates , 57% ( 49/86 ) of called variants were confirmed by Sanger sequencing , comprising 59% ( 10/17 ) of coding variants and 57% ( 39/69 ) of non-coding variants ( Table S5 ) , demonstrating once again the greater accuracy of variant calling in IBD regions compared to non-IBD regions at 4× coverage per mouse . To explore how our computationally efficient and rapid route from phenotypes to candidate genes could be applied to a large-scale ENU program , we proceeded to model the inheritance of mutations within a typical ENU pedigree . First we asked how frequently a hypothetical fully penetrant ENU mutation causing a screened phenotype would be observed among the G3 mice in the proposed breeding strategy . We assumed that a G1 pair could give rise to 4 stable G2 pairs , each generating 3 litters with a conservative estimate of 4 live mice per litter . A single pedigree would then generate 48 G3s for screening ( Figure 1A ) . Using a probabilistic model incorporating all the possible inheritance patterns ( Materials and Methods ) , we calculated that 51% of ENU mutations present in the founder mice occur 3 times or more as homozygous within the set of G3 mice , and 62% occur twice or more . 99 . 6% of mutations would be present at least once as a single allele in the 48 G3 mice , 98% at least 6 times and 95% at least 9 times ( Figure 5A ) . Next , we asked whether , under the proposed strategy , the number of non-causative candidate mutations would be sufficiently small to efficiently exclude these mutations . The number of candidate mutations IBD in affected G3 mice from a single pedigree can be estimated as a function of the number of affected G3s sequenced , the empirical ENU mutation frequency of 1 . 54 mutations ( Figure 3B ) , the relatedness of the mice and the proportion of mutations that affect protein sense . We found on average 1 . 4% ( 9 . 3/647 . 3 ) homozygous mutations in each G3 mouse lie in exons or splice sites; 73% of this subset caused missense or splice site mutations; we identified no homozygous stop mutations in ENU16CH17a . Thus we derive that 1 . 05% of mutations affect protein sense . Using these parameters we calculate that sequencing 3 or 6 mice will reduce the number of candidates to 1 homozygote or 2–3 heterozygote mutations , respectively ( Figure 5B and Materials and Methods ) . This model is consistent with the empirical data from the ENU16CH17a pedigree .
The importance of this study is in showing how low coverage WGS of multiple mice with a phenotype can identify causative ENU mutations without the need for out-crossing , or knowledge of dominant or recessive inheritance . This strategy simultaneously generates linkage maps and identifies the shared mutations with a high degree of confidence . The advantage of our linkage-based approach is that ENU-induced mutations from multiple affected mice can be used to track the IBD regions and then isolate the causative mutation . Linkage analysis using exome sequencing has been effectively harnessed for the study of human pedigrees . Variations on the Lander-Green method , developed for array data , incorporate knowledge of the population allele frequencies of HapMap SNPs [29] , [30] , [31] . Alternative approaches use other hidden Markov models ( HMM ) to identify regions that are IBD and common to autosomal recessive phenotypes [32] , [33] . Reducing the search space from the whole genome to the exome significantly reduces the number of informative variants [34] , however this is typically several orders of magnitude larger than the number of ENU-induced exonic variants , e . g . 6000 to 8000 exonic HapMap variants per individual [29] , [30] compared to 74 exonic variants in our ENU pedigree . The low frequency of ENU coding variants does not permit fine scale linkage analysis based only on exonic variants . Exome sequencing also precludes the detection of regulatory mutations [35] , and the inefficiencies of capture have resulted in a failure to find the causative mutant in one in five ENU pedigrees , even for recessive traits [8] , [36] . The ability to detect IBD regions using low coverage and the falling costs of next generation sequencing make our WGS method increasingly cost effective . Knowledge of the ENU mutation frequency allows us to model an efficient sequencing strategy . Our data show that sequencing 3 affected G3 mice with a recessive trait or 6 mice with a dominant trait would yield on average 1 or 2 candidate IBD mutations . The Lander-Green algorithm on which our IBD analysis is based , scales exponentially with the number of individuals in the pedigree , but remains computationally feasible with a pedigree of individuals and founders where [31] , [37] . The algorithm would accommodate further refinement to take into account the known characteristics of ENU mutations ( Figure 3D ) . By generating haplotype data for many ENU pedigrees , our approach will also eventually lead to a fine scale map of active recombination sites in the mouse , which , unlike existing maps based on recombinations that arose historically between outbred strains of mice [38] or more recently between intercrossed inbred strains [39] , is unbiased by selection or strain differences . Such a map could then be used to optimize the performance of our Lander-Green based algorithm . We believe that the adoption of our approach by large-scale ENU programs will lead to a substantial increase in the productivity of the programs , advancing our understanding of gene-function and the mechanisms of genetic disease . Our approach will reduce the burden in animal costs and allow post-mortem screens , with increased sophistication and accuracy in a broader range of tissues . With the rapidly falling costs of WGS we can envisage a future in which all G3 ENU mice are sequenced to a depth sufficient to identify and segregate all their mutations , creating a rich dataset of allelic variation and corresponding phenotypic information , including linkage data for non-coding mutations with measurable effects . This could be achieved accurately with 4–5× sequencing due to the increased power to impute genotypes . This database could be mined for associations across pedigrees , including the detection of subtle phenotypes .
ENU16CH17aENU treated B6 mice were generated and screened at the Australian Phenomics Facility , The Australian National University , Canberra [40] . Male founder mice for each pedigree , 8–15 weeks old , were treated three times 1 week apart with 90–100 mg/kg N-ethyl-N-nitrosourea ( Sigma ) prepared in 10% ethanol , citrate buffer ( pH 5 . 0 ) . After 8 weeks , the treated mice were mated with B6 females . Individual G1 progeny were intercrossed to generated G2 pairs . Phenotypic screening of G3 mice included flow cytometry of peripheral lymphoid cells ( Figure 1A ) . We obtained tail DNA from 3 affected G3 siblings from the ENU16CH17a pedigree . We performed DNA sequencing on an Illumina HiSeq 2000 machine , using two lanes per mouse . 100 bp paired-end reads were generated . We mapped reads to the mouse reference genome MGSCv37 using Stampy [41] with BWA settings . 94 . 5% of genome was covered at least once , mean coverage across the genome 24 fold per mouse . An in-house variant caller Platypus was used , version 0 . 1 . 8 ( www . well . ox . ac . uk/platypus A . J . R , Mathieson I , G . L , McVean G , ( 2012 ) ) . We annotated the variants using Annovar [42] with Ensembl ( release 64 ) gene annotation . Functional predictions were made using Polyphen-2 using probabilistic classifications based on a model trained with the HumVar dataset , tailored for detection of Mendelian disease caused by mutations with large effects . Although the training dataset consists of human disease causing mutations , the modeling is based on sequence and structural features applicable across species [13] and higher Polyphen-2 scoring has been shown to correlate with damaging murine ENU mutations [36] . The mutation in ENU16CH17a was confirmed independently using exome sequencing [36] . We developed a pipeline to filter the variant calls . As a first step we eliminated variants previously observed in two or more other ENU pedigrees using a variant union file ( Figure S2 ) , since these could be assumed to be due to systematic error , for example in repetitive regions with mismapping , or true non-ENU variation from the reference mouse sequence . To create this union file of shared , and thus non-ENU , variation , Platypus was used to call variants from mice from 9 different ENU pedigrees simultaneously . Thus at each variant locus a genotype and genotype likelihood was assigned for all mice . The nine pedigrees included 6 from the Australian Phenomics Facility at the Australian National University , 2 from the MRC Harwell Centre for Mouse Genetics and one from the Beutler Group at The Scripps Research Institute . The Harwell pedigrees and one of the ANU mice were on a mixed strain background . All other mice were on a straight B6 background ( as was ENU16CH17a ) . We identified 7 , 624 , 313 unfiltered calls amongst these 9 pedigrees from 3 centers ( Figure S2A ) . In order to exclude ENU variation we removed variants observed in only one pedigree . 83 . 6% ( 6 , 371 , 574/7 , 624 , 313 ) of raw variant calls were shared by more than one pedigree ( Figure S2B ) . Within this dataset of shared variants there were 6 , 371 , 548 unique genomic positions , 81 . 7% were SNPs ( 5 , 203 , 298 ) and 18 . 3% ( 1 , 168 , 250 ) indels ) . transition∶transversion ratio among SNPs was 1 . 9% ( 3 , 394 , 771 transitions , 1 , 805 , 983 transversions , 2 , 544 have more than 2 alleles ) . To examine the proportion of this shared variation attributable to B6 reference strain mice , we excluded shared variants exclusively observed in mixed strain mice . The resultant shared variants are observed in at least 2 pedigrees , including at least one fully B6 pedigree . Since all pedigrees included from MRC Harwell are mixed strain , the MRC Harwell variants fully overlap the other laboratories ( Figure S2C ) . The large majority ( 91 . 8% ) of non-ENU variation in B6 mice is not laboratory specific , suggesting isolated genetic drift within individual colonies accounts for little of the observed variation . 24 . 1% of the shared dataset were present in dbSNP . After filtering non-ENU variation from the ENU16CH17a calls , we removed any additional known variants from dbSNP ( version 128 ) and filtered out remaining calls that were clustered closely together with a threshold of less than 1 , 000 bp . The variants were then filtered for Phred based quality score assigned by the variant caller , removal of calls that failed Platypus allele bias or strand bias filters , removal of variants with a high local frequency of bad reads , filters for homopolymers and repetitive sequence , including di-nucleotide repeats of 20 bp or more . Indels were removed as ENU overwhelmingly causes point mutations . Finally loci with coverage in the upper or lower 1% of the coverage distribution were excluded . All coverage distributions were measured with BEDTools [43] . The Lander-Green algorithm [10] uses a combination of genotype information from informative markers ( here , the SNP genotypes called from the next-generation sequencing data ) , and knowledge of local recombination rates to determine the ancestral haplotypes in specific genomic intervals , and the locations at which recombinations and therefore transitions between inheritance state vectors occur . The pedigree used in this experiment ( Figure 1A ) can be considered to originate with the G1 pair of common ancestors who each carry a combination of the haplotypes inherited from the G0 mice , these being ENU1 , ENU2 , WT1 and WT2 . Thus for the purpose of the algorithm the pedigree consists of 5 non-originals – that is 5 individual mice with parents in the pedigree . The Lander-Green algorithm represents the ancestral haplotypes of the 5 non-originals as a state vector with 10 binary co-ordinates , representing the 5 individual mice , arising from 10 gametes . Within each inheritance vector a 0 coordinate indicates a gamete carrying grand-paternal DNA at a locus , and 1 indicates grand-maternal inheritance . These are arbitrarily phased . There are possible state vectors . The standard HMM machinery is well documented [44] . A HMM has two main components which are model-specific: these are the state transition matrix , which specifies the probabilities of transitions between any two model states , and the likelihood , which is the probability of the observed data given a particular model state . The Lander-Green algorithm uses a state transition matrix based on the recombination rate , which encodes the probabilities of transitions between any of the ancestral state vectors , based on the number of recombinations required for the transition . In our implementation we use a recombination map [39] to compute average local recombination rates across the genome . We do not store the whole transition matrix in memory , but compute matrix elements on demand . This is straight-forward , as all matrix entries can be expressed as powers of the recombination rate and one minus the recombination rate . This vastly reduces the memory requirements for the algorithm , which are now linear in the number of state vectors rather than quadratic . For each G3 , a state vector determines which two ancestral haplotypes that make up G1 , make up the local genotype , e . g . ( ENU1 , WT2 ) . We compute a probability for the observed SNP genotypes , given each ancestral state vector , in 100 kb windows across each chromosome . This probability has two components: a prior probability of observing a SNP in the given window , and genotype likelihoods computed by the variant caller from the sequence data . We assume fixed priors of observing a SNP in ENU haplotypes ( ) and WT haplotypes ( ) . The IBD regions inferred by the algorithm were relatively insensitive to changes in these priors ( Figure S3B ) . The likelihood for a particular state vector is the sum over all possible combinations of the SNP genotypes ( 0/0 , 1/0 , 1/1 ) for the 3 mice , of the product of the SNP priors and the relevant genotype likelihoods for the 3 G3 mice . This incorporates the dependent relationships between the mice . In the case of multiple SNPs occurring in the same window , we assume that the SNPs are independent , and the likelihood for all mice in the window is the product of the likelihoods across all SNPs . Using the genotype likelihoods from the caller allows us to accommodate errors in the WGS data; a modification to the conventional Lander-Green algorithm that has been used to infer IBD in array data . [45] . Due to the paucity of polymorphic sites in the in-bred B6 mouse , there are many 100 kb windows which contain no SNPs . If no SNPs were called in a window , the most likely explanation is that no SNPs were present . In a small fraction of cases , a real SNP will be missed due to low coverage or variant-calling errors . To deal with windows that contain no SNPs , we supply a set of likelihoods weighted towards 0/0 . Specifically we assume a 1/10 probability that a heterozygous SNP was missed , and a 1/100 probability that a homozygous SNP was missed . Finally , we use the forward backward algorithm [44] to compute the posterior probability of each state vector at each window . We select the state vector with the highest posterior in each window to construct the sequence of most probable inheritance states across each chromosome . This information is shown in Figure 2A . The program was coded in Python ( http://www . python . org ) and Cython ( http://cython . org ) . The code is freely downloadable from http://www . well . ox . ac . uk/lgenu . Graphical plots of genotypes were generated using matplotlib ( http://matplotlib . sourceforge . net ) . Within our three generation pedigree , frequent recombinations , particularly from state A to state B and back to state A within a small genomic interval are likely to be artifactual . Therefore we performed a smoothing step on the output inheritance states , such that recombinations from state A to state B and back to state A within 1 Mb , within an allele , are corrected to state A . In reporting the results we define IBD heterozygous throughout to refer to regions or variant sets in which all sequenced individuals share at least one allele from the same ENU founder but are not IBD homozygous . Our dataset includes on average 647 mutations per mouse in homozygous regions spanning over 1 , 000 Mb across 3 mice ( mean 405 Mb , 15 . 3% of the genome , per mouse ) . Sanger sequencing of the candidate mutations confirms the reliability of our filtered call set ( Table S1 ) . By observing the frequency of these mutations and subtracting the low background frequency from homozygous WT regions , we estimated the ENU mutation frequency , and used the homozygous ENU region variants to examine the base preference of ENU mutations . Comparison of the transition transversion ratio was made with all the variants in the Centre for Genome Dynamics Mouse SNP Database ( http://cgd . jax . org/cgdsnpdb ) on 8 . 5 . 12 . This database includes over 66 million SNPs from 136 inbred laboratory mouse strains predominantly imputed from the mouse diversity array [46] , and is representative of the characteristics of naturally occurring ( non-ENU ) SNPs in an inbred mouse . We generated random subsets of reads from the ENU16CH17a bam file using a script that utilizes pysam , a SAMtools [47] interface for Python . For details see http://code . google . com/p/pysam/ . We called and filtered variants on the down-sampled bam files using the same pipeline described above . Comparisons of IBD regions and variant calls between the lower coverage datasets and the full ( 24× ) coverage data were made using BEDTools [43] to measure intersections . The intersections of IBD regions were analyzed by a per base pair comparison . To examine whether an IBD method performs better than a simple per SNP approach to detect shared variation in the 3 mice attributable to ENU we generated a comparison set of shared variants at each simulated coverage depth by selecting variants that were observed in all 3 mice . In exactly the same way as with the IBD variants we included SNPs in which there was at least one homozygous or heterozygous mouse and the remaining 0 , 1 or 2 mice had no genotype information ( denoted in the vcf file ) or a reference ( ) genotype call with at least one variant call and less than 5 supporting reference reads . To examine the distribution between missense and splice mutations we looked at the larger dataset of heterozygous mutations . Across our three mice , 86% ( 21 . 7/25 ) of potentially damaging mutations were missense mutations , 6% ( 1 . 3/25 ) were nonsense mutations and 8% ( 2/25 ) were in splice sites . A large database ( http://mutagenetix . utsouthwestern . edu/ on 8 . 5 . 12 ) of over 5 , 000 incidental ENU mutations with no observed deleterious effects , identified in the course of next generation sequencing ( Applied Biosystems SOLiD ) , reports 85 . 9% missense , 4 . 4% nonsense and 9 . 7% splicing mutations , and agrees broadly with our findings . To model the segregation of mutations within a pedigree , we calculated the probability of a mutation being inherited by a G3 under the three possible situations where the G2 parents carry 0 , 1 or 2 copies of the mutation ( in 25% , 50% and 25% cases respectively; individual G2 mice are not homozygous for any ENU mutation ) . We denote the chance of inheriting the mutation at G3 under each of these situations with a given zygosity ( homozygous or heterozygous ) at G3 as , , and . Clearly some of these probabilities will be 0 . Conditional on the number of mutations carried by the G2 parents , the number of G3 offspring inheriting the mutation with the required zygosity can be modeled using a binomial distribution . For a given G2 parent pair , we denote this distribution by . Assuming that each G2 pair produces 3 litters of 4 live mice , this distribution is given by ( 1 ) Here or 12 choose denotes the binomial co-efficient indexed by 12 and . To estimate the probability of G3 carrying the mutation across the 48 G3 from 4 G2 pairs , we can convolute across all combinations of mice that together transmit precisely mutations with the required zygosity from G2 pairs , such that ( 2 ) We considered two situations: one of a recessive mutation , in which a G3 has two alleles from parents that are heterozygous for the mutant; and the situation of a dominant mutation , where we considered that homozygotes may also have the phenotype and may be indistinguishable from heterozygotes . In this way it is possible to calculate the probability of any recessive or dominant mutation carried by a founder occurring M times in the G3 mice . The results are presented in Figure 5A . We modeled the proportion of the genome expected to be IBD in sequenced mice without accounting for linkage to the causative mutation . We calculated the probability of G3 carrying a shared ancestral haplotype at any locus as described above in equation 2 , for each between and 48 ( the modeled number of G3 mice ) , and calculated a probability of picking mice sharing such a locus by chance from a pool of mice sharing the locus and individuals not carrying the locus . Since each mouse can only be picked once this corresponds to a hypergeometric distribution . We then summed over the product of this and the for each between and 48 to get the overall probability , , of any unlinked locus being observed in all the affected sequenced mice . ( 3 ) is the proportion of the genome expected to be IBD for a specified number of sequenced mice . We used our knowledge of the ENU mutation rate ( 1 . 54 mutations ) , and the fraction of variation affecting protein sense ( 1 . 05% missense , nonsense or splicing ) , to estimate the number of homozygous or heterozygous candidate mutations shared by affected sequenced mice . We add one mutation to model the causative mutation which is always present . In phenotypically affected mice a region from the ENU founder persists around the causative mutation due to linkage , and this adds another fraction of the genome , where c is the fractional size of the chromosome , m is the number of meioses per G3 mouse , and k is the number of G3 mice . This approximates to a further 0 . 7 mutations or candidate coding mutations . Since this is negligible we simply approximate to 1 additional mutation . The results are presented in Figure 5B . We amplified the 28 candidate mutations with two rounds of PCR from genomic DNA using internal and external fully nested primers ( Table S3 ) and then amplified with Big Dye ( Applied Biosystems Ltd ) before sequencing on an Applied Biosystems 3720xl machine . All nested sequencing reactions were run in duplicate to check for PCR error . We carried out Sanger sequencing to validate shared variants from the 3 mice sequenced at low coverage from a second pedigree using primers shown in Table S4 and Table S5 . Analysis of means was performed using the Graphpad Prism 5 package . All other analyses were written in custom scripts and described in the Materials and Methods . All animal experiments were approved by local and national ethical review , including the Australian National University Animal Ethics and Experimentation Committee and the Oxford University Local Ethical Review Committee and UK Home office License PPL 30/2455 . | Damaging mutations in single genes are an important source of information about the causes of disease , including more complex genetic disease; but these single gene disorders are typically rare in humans . An important strategy for identifying new disease mechanisms is to introduce multiple random mutations in mice and test the mice for biological differences; these mice act as models of human disease . However , discovering the disease-causing mutation is time-consuming and complex , requiring further generations of breeding . In this study we demonstrate a method that overcomes these problems by sequencing the entire genomes of multiple mice that have inherited a disease-causing mutation from a common ancestor . We use an algorithm that uses knowledge of all the mutations carried by the sequenced mice to identify the regions of the genome and mutations that are common to all the mice . Using this method we can rapidly link biological traits to genetic mutations . In contrast to current approaches , our strategy does not require large amounts of breeding , and it permits more accurate measurement of a wider range of traits; consequently its introduction will significantly reduce the number of mice required in the future , increase the number of traits that can be detected , and accelerate the discovery of new pathways and gene functions relevant to human diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"genetics",
"genetic",
"mutation",
"population",
"genetics",
"immunology",
"gene",
"function",
"animal",
"models",
"mutation",
"model",
"organisms",
"linkage",
"(genetics)",
"sequence",
"analysis",
"genetic",
"polymorphism",
"biology",
"mouse",
"genotypes",
"m... | 2013 | Unlocking the Bottleneck in Forward Genetics Using Whole-Genome Sequencing and Identity by Descent to Isolate Causative Mutations |
Primary Ciliary Dyskinesia is a heterogeneous genetic disease that is characterized by cilia dysfunction of the epithelial cells lining the respiratory tracts , resulting in recurrent respiratory tract infections . Despite lifelong physiological therapy and antibiotics , the lungs of affected patients are progressively destroyed , leading to respiratory insufficiency . Recessive mutations in Dynein Axonemal Intermediate chain type 1 ( DNAI1 ) gene have been described in 10% of cases of Primary Ciliary Dyskinesia . Our goal was to restore normal ciliary beating in DNAI1–deficient human airway epithelial cells . A lentiviral vector based on Simian Immunodeficiency Virus pseudotyped with Vesicular Stomatitis Virus Glycoprotein was used to transduce cultured human airway epithelial cells with a cDNA of DNAI1 driven by the Elongation Factor 1 promoter . Transcription and translation of the transduced gene were tested by RT–PCR and western blot , respectively . Human airway epithelial cells that were DNAI1–deficient due to compound heterozygous mutations , and consequently had immotile cilia and no outer dynein arm , were transduced by the lentivirus . Cilia beating was recorded and electron microscopy of the cilia was performed . Transcription and translation of the transduced DNAI1 gene were detected in human cells treated with the lentivirus . In addition , immotile cilia recovered a normal beat and outer dynein arms reappeared . We demonstrated that it is possible to obtain a normalization of ciliary beat frequency of deficient human airway epithelial cells by using a lentivirus to transduce cells with the therapeutic gene . This preliminary step constitutes a conceptual proof that is indispensable in the perspective of Primary Ciliary Dyskinesia's in vivo gene therapy . This is the first time that recovery of cilia beating is demonstrated in this disease .
Primary Ciliary Dyskinesia ( PCD , OMIM #242650 ) is an inherited disease mainly characterized by dysfunction of airways' motile cilia . The prevalence is approximately 1 in 12 , 000–20 , 000 [1]–[3] . About 50% of patients affected by PCD have a situs inversus which results from monocilia dysfunction at the embryonic node [4] . This association is referred to as Kartagener's syndrome ( OMIM #244400 ) [5] . PCD causes chronic sinus and bronchial respiratory infections that begin early in life , leading to nasal polyps and bronchiectasis . Males are frequently sterile due to dysfunctional spermatozoa flagella [6] . Other symptoms can also be associated with PCD like hydrocephalus , anosmia , retinitis pigmentosa and congenital heart diseases [7]–[10] . The disorder is genetically heterogeneous and in most cases , inheritance is autosomal recessive but X-linked inheritance patterns were also described [11] . Several loci and some genes have been identified , as DNAI1 ( Ensembl ENSG00000122735 ) , DNAH5 ( ENSG00000039139 ) , DNAH11 ( ENSG00000105877 ) , RPGR ( ENSG00000156313 ) , TXNDC3 ( ENSG00000086288 ) , OFD1 ( ENSG00000046651 ) , DNAI2 ( ENSG00000171595 ) and KTU ( alias C14orf104 ) ( OTTHUMG00000152331 ) genes [10] , [12]–[19] . The first gene described to be responsible for PCD and Kartagener syndrome was DNAI1 gene [18] , [20] . Eighteen mutations in DNAI1 gene were reported , and Zariwala et al . evidenced a founder effect for the most frequent mutation ( c . 48+2_48+3insT ) [21] . Moreover , the authors estimated that mutations in DNAI1 gene represent about 10% of PCD cases . DNAI1 encodes an axonemal dynein intermediate chain , a component of the outer dynein arm ( ODA ) . Dyneins are molecular motors which produce energy for microtubules doublets sliding in the axoneme . To date , no etiologic treatment of PCD is available and on the long range , PCD leads to respiratory insufficiency and lung transplant . We hypothesized that gene therapy could restore ciliary function in DNAI1-mutated airway epithelial cells to prevent patients from infectious complications . To introduce genetic material into cells we focused on lentiviral gene transfer because lentivirus has the property to integrate its genetic material into host cell genome even in non-replicating cell [22] . Moreover , lentivirus is weakly immunogenic unlike recombinant adenovirus which efficiency was reported to decrease after several administrations in a clinical study of patients suffering from cystic fibrosis [23] . Lentiviral-derived vectors used in gene therapy were principally based on SIV ( simian immunodeficiency virus ) or HIV ( human immunodeficiency virus ) . For this latter one , gene transfer efficiency into bronchial epithelium in mice was already demonstrated [24] . We decided to modify a SIV-based vector , previously described by Negre et al . to efficiently transduce mature human dendritic cells [25] , [26] , and to transduce human airway epithelial cells ( HAECs ) cultured as described by Jorissen et al . [27] . First , we showed here that normal HAECs were efficiently transduced by SIV-based vector containing eGFP gene . Then , we validated lentiviral vectors' constructions containing DNAI1 cDNA sequence and showed that transduced DNAI1 is transcribed and expressed . Finally , we demonstrated that transduction of DNAI1-mutated HAECs with wild-type DNAI1 can restore ciliary beating and that ODA are binding again to microtubules .
To estimate whether HAECs could be transduced by a SIV-based lentivirus pseudotyped with VSV-G , normal HAECs were infected with pGFP a vector containing eGFP as a reporter gene in a variety of conditions [25] , [26] . Three parameters were investigated: ( 1 ) the multiplicity of infection ( MOI ) , ( 2 ) the moment of infection and ( 3 ) whether using a polycation , Polybrene , or not . Two different moments of infection during Jorissen's culture were tested: at J+1 , cells were ciliated and in suspension or at J+3 , cells were de-differentiated and adherent . Two days post-infection , reporter gene expression was analyzed by FACS ( Figure 1 ) . In any of the selected conditions , HAECs were transduced but the proportion of transduced cells seemed to be dependant on MOI irrespective of the other parameters , and apparently higher at MOI 75 . Then , transduction of cells infected at J+1 seemed more efficient compared to cells infected at J+3 . At MOI 75 , approximately 38% of cells infected at J+1 were transduced versus 20% for cells infected at J+3 . Moreover , these results seemed to be improved by the use of Polybrene . Finally , at MOI 75 , transduction efficiency was quantified at about 38% for cells infected at J+1 without Polybrene compared to approximately 50% for cells infected at J+1 with the use of Polybrene . These experiments which were not repeated , demonstrated that HAECs could be transduced in a variety of conditions and that J+1 with Polybrene at a MOI of 75 were presumably the best experimental conditions for HAECs infection . Therefore , we selected these conditions for further experiments . Full-length DNAI1 cDNA was cloned into the lentiviral vector in place of eGFP gene . Two constructions were obtained using BamH I and Xho I restriction sites ( pK-DNAI1 ) or Nco I and Xho I ( pK+DNAI1 ) , and resulted respectively in an intact DNAI1 cDNA associated with a modified Kozak sequence which could prevent normal translation , or an intact Kozak sequence associated with a modified DNAI1 cDNA which could result in a dysfunctional protein ( Figure 2A ) . Then , to differentiate endogene from exogene DNAI1 , a hemagglutinin tag ( HA ) was added at the 3′ side of DNAI1 cDNA sequence in each plasmid which resulted in two additional vectors: pK-HA or pK+HA ( Figure 2B ) . Normal HAECs were transduced at J+1 with Polybrene at a MOI of 75 since these conditions gave satisfactory results as evaluated with eGFP . First , mRNA extracts from transduced HAECs were controlled by PCR with alpha-tubulin specific primers for absence of genomic DNA ( gDNA ) contamination ( not shown ) . Second , we confirmed that non-ciliated HAECs ( NC ) do not express DNAI1 because DNAI1 specific RT primer ( P5 ) led to an absence of amplification by contrast to re-ciliated HAECs ( RC ) template ( Figure 3 , lanes 1 and 2 ) . Third , DNAI1 mRNA was not amplified using a HA-tag RT specific primer ( HA ) from non-infected re-ciliated HAECs ( Figure 3 , lane 3 ) . To test DNAI1 transcription from lentiviral vectors , HA-tagged DNAI1 gene transcription was revealed by RT-PCR using HA-DNAI1 specific primers . Re-ciliated HAECs infected by particles containing HA-tagged DNAI1 with either an exact Kozak sequence ( pK+HA ) or a modified Kozak sequence ( pK-HA ) both transcribed HA-tagged DNAI1 ( Figure 3 , lanes 4 and 5 ) . Non-ciliated HAECs infected by pK-HA particles also expressed HA-tagged DNAI1 ( Figure 3 , lane 6 ) . In conclusion , DNAI1 transcription is efficient from vectors containing a conserved or a modified Kozak sequence and transcription of the transduced DNAI1 cDNA driven by Elongation Factor-1 promoter is active in non-ciliated and re-ciliated HAECs . To detect DNAI1 protein translation from lentiviral vectors , 293Bosc cell line which naturally does not express DNAI1 was transfected by each lentiviral vector using ExGen500 . DNAI1 protein was detected by Western blot with specific anti-DNAI1 antibodies after 293Bosc cells' transfection with the Kozak modified ( pK-DNAI1 ) or Kozak conserved ( pK+DNAI1 ) DNAI1 sequence ( Figure 4 , lane 1 and 2 ) . DNAI1 protein was also detected after 293Bosc cells' transfection with the vectors containing HA appended to the 3′ end of DNAI1 sequence ( Figure 4 , lane 3: modified Kozak and lane 5: conserved Kozak sequence ) . By contrast , no DNAI1 protein could be detected by Western blot in protein lysate of non-transfected 293Bosc cells with anti-DNAI1 antibodies ( Figure 4 , lane 4 ) . Moreover , immunofluorescence assays showed that in pK-HA transfected 293Bosc cells , cytoplasmic HA-tagged DNAI1 proteins could be specifically detected by HA antibodies ( Figure S1 ) . In the second set of experiments , transduction was performed on DNAI1-mutated HAEC at J+1 and cells were cultured to generate ciliated vesicles . DNAI1-mutated HAECs were transduced with pK-DNAI1 or pK-HA to determine if immotile cilia might recover a beat . The same protocol of transduction was used as for normal HAECs . We confirmed DNAI1-mutated HAECs transduction efficiency with pGFP vector , as de-differentiated cells at J+3 and re-ciliated cells at J'+17 ( 17 days post-collagen digestion ) expressed GFP protein ( Figure S2 ) . After re-differentiation , GFP transduced HAECs were covered with cilia ( Figure 5A-A ) but these cilia were immotile . The variation of optic signal along a line crossing the cilia during 400 msec does not show any movement ( Figure 5A-B ) . This immotily is also visible on video recording ( Video S1 ) . By contrast , ciliary beating was recorded on DNAI1-mutated HAECs transduced with either pK-DNAI1 or pK-HA vector ( Figure 5A-C ) . This beat is demonstrated by recording the variation of optic signal along a line crossing cilia . Waves are clearly visible ( Figure 5A-D ) that evidenced the periodic beat of cilia . This active beat is also visible on video recording ( Video S2 ) . At J'+30 , a ciliary beat frequency ( CBF ) was measured for pK-DNAI1 and pK-HA transduced cells . CBF of HAECs with beating cilia were 9 . 95±1 . 23 Hz and 11 . 31±0 . 85 Hz , for pK-DNAI1 and pK-HA treated cells , respectively . These values fall into the range of control HAECs ( from 7 to 11 Hz ) . Cilia length in DNAI1-mutated HAECs was estimated at 6 µm in all cases and did not depend on DNAI1 treatment and cilia beating . In DNAI1-mutated HAECs' axoneme , ultrastructure analysis by TEM showed that ODA were absent or shorter than in normal HAECs ( Figure 5B-B ) . TEM on axonemes of DNAI1-treated cells were analyzed and some cilia had normal amounts of ODA , while in other cilia the ODA were partially absent ( Figure 5B-C and 5B-D ) . The average number of ODA per axoneme was 3 . 29±1 . 53 in pGFP infected DNAI1-mutated HAECs . ODA increased significantly to 5 . 67±1 . 83 ( p<0 . 0001 ) and 5 . 73±2 . 10 ( p<0 . 002 ) in DNAI1-mutated HAECs treated with pK-DNAI1 and pK-HA , respectively . By contrast , there was no significant difference between DNAI1-mutated HAECs treated by pK-DNAI1 or pK-HA . The distribution of the number of ODA per axoneme presented a single peak in pGFP infected DNAI1-mutated HAECs but two peaks in pK-DNAI1 and pK-HA treated cells ( Figure 6 ) . IDA analysis showed no difference between control and DNAI1-treated cells ( data not shown ) .
In the present study , we demonstrated that a SIV-based vector pseudotyped with VSV-G protein , previously described to efficiently transduce human dendritic cells [25] , [26] , could also efficiently infect normal cultured HAECs [27] . Previous studies showed that Murine Leukemia virus ( MuLV ) [28] and Feline Immunodeficiency Virus pseudotyped with VSV-G envelope [29] could transduce HAECs in culture . Though , this transduction was only possible from the basolateral surface of polarized HAECs . We did not know whether a Simian Immunodeficiency Virus would transduce although we could foresee that transduction would be more efficient from the basolateral surface of HAECs since it was pseudotyped with VSV-G . We did not know also whether de-differentiated cells would be transduced . It appeared that whatever the conditions used in this study , a certain proportion of cultured HAECs were infected . We then selected the conditions that seemed to provide the higher percentage of infected cells in culture . In any case , these conditions are strikingly different from in vivo situation where HAECs are a component of a complex epithelium recovered by a thin layer of mucus . Moreover , Polybrene which seemed to improve viral infection in cultured cells cannot be used in vivo due to side-effects . In a next step , we demonstrated transcription and translation of the transduced DNAI1 cDNA . Finally , we showed that DNAI1-mutated HAECs treated by DNAI1 gene transfer recovered ciliary beating whereas GFP-treated cells' cilia remained immotile , thus supporting our concept that PCD gene therapy was possible using DNAI1 gene . This conceptual proof is essential in the perspective of a human gene therapy . Since it was impossible to obtain a vector containing a conserved Kozak sequence associated with a normal DNAI1 cDNA sequence , we decided to construct two vectors: one with a modified Kozak sequence upstream a conserved DNAI1 cDNA sequence ( pK-DNAI1 ) , and another one with a conserved Kozak sequence but associated with a modified DNAI1 cDNA sequence ( pK+DNAI1 ) . However , both vectors turned out to be equally efficient in terms of transcription and translation . Moreover , the addition of a HA tag at the 3′ end did not alter the transcription neither the translation of both vectors . Consequently , we pursued the infection experiments on DNAI1-mutated HAECs with the Kozak modified/DNAI1 conserved pair of vectors ( with and without HA tag ) . We were unable to detect endogenous DNAI1 protein and DNAI1 HA-tagged protein on Western blot of cultured normal HAECs presumably because the amount of extracted proteins from cultured HAECs was too low . However , translation from the various vectors was confirmed in transfected 293Bosc cells . We also failed to label HA-tagged DNAI1 protein in the cilia of infected HAECs . However , the observation of the rescue of ciliary beating after DNAI1 treatment of DNAI1-mutated HAECs with either one of the selected vectors supported the idea that normal DNAI1 proteins , HA-tagged or not , were efficiently localized to cilia . This functional evidence was reinforced by ultrastructural analysis showing an increased number of ODA in axonemes with tagged and untagged DNAI1 proteins . Beside these observations , no difference in the number of IDA was detected between control and DNAI1-treated HAECs , in agreement with the knowledge that DNAI1 proteins specifically belong to ODA . In DNAI1-mutated HAECs treated by DNAI1 , a partially uncoordinated ciliary beating was observed , probably because as we demonstrated in early experiments , about half of infected cells were not transduced . As a consequence , HAECs aggregates are mosaics composed of cells with immotile cilia and other cells with beating cilia . The number of ODA per axoneme in treated DNAI1-mutated HAECs is lower than normal because these values reflected a mix population of HAECs: efficiently transduced cells and untransduced cells , as shown in the bimodal distribution of ODA per axoneme in Figure 6 . In this respect , our results are consistent with the recessive inheritance type of PCD disease for which a normal copy of the DNAI1 gene is sufficient to an absence of symptoms . In order to improve the number of rescued HAECs , we need to improve transduction efficiency , but respiratory epithelium is not exclusively composed of ciliated cells and at least seven other cell types are present [30] . To produce viral particles which would preferentially target human airway ciliated cells , the VSV-G large cell tropism protein could be replaced by another envelope protein from one of the various viruses known to infect airway epithelium , ie . coronavirus or Influenza virus [31] , [32] . Influenza virus receptors , as α2 , 3-linked sialic acid receptors , were described to be specifically localized on ciliated airway and type II alveolar epithelial cells in both mouse and human respiratory epithelium , and on a subpopulation of basal cells in humans [32] . By contrast , α2 , 6-linked sialic acid receptors are expressed on human ciliated and goblet cells but not in the mouse . Therefore , it should be interesting to evaluate on our current culture model , the use of hemagglutinin ( HA ) from selected Influenza viruses which preferentially bind to ciliated cells . Finally , Szecsi et al . demonstrated that HA envelope protein of H7N1 and H5N1 avian viruses was successfully associated with retroviral-based vectors resulting in higher titres of infected cells than with VSV-G [33] . Airway ciliated cells are differentiated cells which do not proliferate and their cycle life span is supposed to be relatively short . In the mouse model , ciliated cells renewal was reported to be provided by Clara cells which are also essential in the alveolar type II cells generation [34] . Nevertheless , Park et al . reported that after an injury event , murine ciliated epithelial cells could transdifferentiate into other cell types and restored the airway epithelium [35] . In the human airway epithelium , Hajj et al . demonstrated that basal cells were able to regenerate as well as to differentiate into secretory and ciliated cells [36] . Targeting these basal cells with a lentiviral vector would result in the stable integration of the therapeutic gene with no need to repeat infections . The specificity of cells' targeting has been recently improved by taking into account the presence of proteases at the cell surface [37] . Thus , cell transduction can be achieved only in cell types that have the capacity to bind viral surface glycoprotein and to cleave the protease substrate because the surface glycoprotein of several enveloped viruses are expressed as precursor proteins that need to be cleaved into two subunits by cellular proteases to promote cell entry . Finally , DNAI1 gene transfer could be applied to other genes causing PCD , as DNAH5 mutations are responsible for at least 28% of PCD cases [21] . However , another vector should be considered for this gene because of size limitation of lentivirus . As an alternative , AAV vectors offer a good safety profile in particular because viral DNA does not integrate into the genome of transfected cells . However , AAV repeated administration triggers anti-viral immune response and its packaging capacity is too limited to hold the DNAH5 gene ( 13 . 9 kb ) . High-capacity episomal vectors offer the possibility to deliver not only a large gene but also the whole genomic locus including regulatory regions which ensures a physiological expression with tissue and time specificity . However , delivery of extrachromosomal DNA with non viral or viral strategies is so far limited to cell lines . This system , tough , is appropriate for long-term treatment as in PCD because chromosome-based episomal systems contain elements enabling the cell to process episome as an additional chromosome with a mitotic stability >99% in cell lines [38] .
The diagnosis of a male patient affected by PCD was based on clinical signs ( recurrent upper and lower respiratory tract infections since early in life , nasal polyps and complete situs inversus – Kartagener syndrome ) . In addition , he was sterile . Electron microscopy of the cilia showed no ODA . Ex vivo culture of epithelial airway cells according to Jorissen's method demonstrated that his cilia were 6 µm-long and totally immotile . Analysis of his DNA demonstrated that he harbored heterozygous compound DNAI1 mutations: c . 48+2_48+3insT and c . 1543G>A [20] . This patient signed an informed consent allowing us to experiment on nasal biopsies which consisted in removing the most anterior part of his middle nasal turbinate on both side . This study complies with the rules of the local ethical committee . Human airway epithelial cells ( HAECs ) from normal subjects were obtained from nasal turbinates which were removed and discarded in the process giving access to the ethmoidal sinus . Patients were operated for tumours located in the ethmoidal region and had no respiratory disease . Cells from control subjects and the patient were grown using the immerged cell culture previously described by Jorissen et al . [27] . Briefly , ciliated cells were isolated and cultured the day following the biopsy ( J+1 ) in collagen-coated flasks to de-differentiate in non-ciliated cells . When they reached 80–90% confluence , collagen was digested ( J' , 7–10 days post-seeding ) and cells were suspended in flasks with rotation to re-differentiate in the form of ciliated vesicles . Cells were infected at J+1 or J+3 . Non-ciliated cells were harvested at J+7 to 10 ( J' ) and re-ciliated cells were fully re-differentiated at J'+28 . To generate DNAI1 cDNA ( AF091619 ) , we extracted total RNA from ciliated HAEC and synthesized the cDNA . A pair of primers was designed to amplify the full-length cDNA . Six supplementary sets of specific primers were used to sequence DNAI1 cDNA . The full-length DNAI1 cDNA PCR product was cloned . Addition of enzyme restriction sites in 5′ and 3′ DNAI1 cDNA sequence was performed using two sets of primers . The BAMHIDNAI1_for and NCOIDNAI1_for primers ( forward ) added BamH I and Nco I restriction sites upstream DNAI1 cDNA sequence , respectively . The XHOIDNAI1_rev primer ( reverse ) added a Xho I restriction site downstream DNAI1 cDNA sequence . Addition of hemagglutinin ( HA ) tag downstream DNAI1 cDNA sequence was performed by PCR using lentiviral vectors containing DNAI1 cDNA ( pK+DNAI1 or pK-DNAI1 ) as template and upDNAI1_for ( forward ) /lowHA_rev ( reverse ) primers . For more information see Text S1 . The lentiviral vector system used in this study was derived from SIV vectors described by Negre et al . [25] , [26] . Five different lentiviral vector constructs were used all under the transcriptional control of the human Elongation Factor-1 promoter ( EFS ) : ( 1 ) pR4SA-EFS-GFP-W ( pGFP ) , ( 2 ) pK-DNAI1 , ( 3 ) pK+DNAI1 , ( 4 ) pK-HA , ( 5 ) pK+HA . The pGFP vector contains the eGFP cDNA sequence . The pK-DNAI1 and pK+DNAI1 vector constructs are essentially the same as the pGFP vector but with the DNAI1 cDNA sequence replacing the eGFP gene sequence , using BamH I/Xho I and Nco I/Xho I sites , respectively ( Figure 2A ) . The pK-DNAI1 vector has a modified Kozak sequence with an intact DNAI1 cDNA sequence whereas pK+DNAI1 has an intact Kozak sequence with a modification at position 4 of the DNAI1 cDNA sequence , resulting in DNAI1 second amino acid modification: isoleucine>valine . The pK-HA and pK+HA vector constructs correspond to pK-DNAI1 and pK+DNAI1 vectors , respectively , with a 3′ DNAI1 cDNA hemagglutinin ( HA ) tag addition , using Blp I/Xho I restriction sites ( Figure 2B ) . Three sets of experiments were carried out . In the first set , normal HAECs were transduced the day of seeding ( J+1 ) on collagen-coated flasks or at J+3 , and the cells were incubated for 24 hours with the lentivirus at MOI ( Multiplicity Of Infection ) of 7 , 35 and 75 in the presence or absence of 6 µg/mL Polybrene ( 1 , 5-dimethyl-1 , 5-diazaundecamethylene polymethobromide , hexadimethrine bromide ) . MOI is the ratio of infectious agents ( e . g . lentivirus ) to infection targets ( e . g . cells ) . Then , medium was completely changed in order to remove debris and inactive lentiviruses . Two days post-infection , reporter gene expression was analyzed by FACS . In the second set of experiments , normal HAECs were transduced or not the day of seeding ( J+1 ) at MOI75 with Polybrene with pK-HA or pK+HA . RNA was extracted before or after re-ciliation . In the third set of experiments , transduction was performed on DNAI1-mutated HAECs at J+1 , at MOI 75 , with the use of Polybrene and cells were cultured to generate ciliated vesicles . Tagged DNAI1 gene expression was revealed by reverse transcription PCR ( RT-PCR ) on infected HAEC . Non-ciliated cells were collected the day of collagen digestion ( J' ) and ciliated cells were collected when they were fully covered by cilia ( J'+28 ) . Poly ( A ) + mRNA was isolated by the Dynabeads Oligo ( dT ) 25 purification kit , according to the manufacturer's protocol ( Dynal Biotech , Norway ) . Transient transfection of 293Bosc cells was performed with each four different plasmids: pK-DNAI1 , pK+DNAI1 , pK-HA or pK+HA . Western blot analysis was carried out according to standard techniques ( Text S1 ) . In order to assess the functional activity of the ciliated DNAI1-mutated cells after lentiviral transduction , video recordings were performed with a ×40 objective lens in the light path at different steps of the culture in flasks , by using an Olympus IX50 inverted phase-contrast microscope . The control eGFP fluorescence was observed with a magnifying digital camera SCION CFW 1308M ( Scion Corporation , Frederick , MD ) and the recovery of ciliary beating was recorded with high speed digital video camera pco . 1200 hs ( PCO , Germany ) . The digital image-sampling rate was software-controllable using CamWare and for all experiments the sampling rate was set at 500 frames per second ( fps ) . For measurements of ciliary beat frequency ( CBF ) , the video images of active ciliated cells were captured with a ×100 oil-immerged objective lens , using Leica DMRXA microscope and pco . 1200 hs camera . A time-motion representation was obtained and analyzed using ImageJ software ( National Institutes of Health , USA ) . Briefly , a line cutting cilia close to their tip was drawn . The “reslice” function of ImageJ was then used to obtain the video signal along this line ( y axis ) during 400 msec ( x axis ) . Finally , the data were expressed as mean±SD from at least three regions of interest . Values for normal HAECs were used as controls . Transmission electron microscopic analyses were processed on ciliated vesicles obtained after HAECs culture , as described by Jorissen et al . [27] . Amounts of ODA were expressed as mean±SD from n axonemes ( n = 17 for pGFP , n = 21 for pK-DNAI1 , n = 11 for pK-HA ) . Values of HAEC infected with pGFP , pK-DNAI1 or pK-HA were statistically analysed by Student t test and the significance was calculated for a two-tailed test . See Text S1 for more details . | This manuscript reports on a successful gene therapy attempt on human airway epithelial cells of a patient suffering from Primary Ciliary Dyskinesia . In this autosomal recessive disease , cilia of the epithelial cells that border the upper and lower respiratory tracks are not functioning . As a result , patients suffer from recurrent airway infections leading progressively to respiratory insufficiency . There is no treatment as of today that could restore normal ciliary beating . In this report , we showed that it is feasible to transfer a therapeutic gene to human airway epithelial cells with a lentivirus . This transferred gene is transcribed and expressed . Moreover , defective cells that had immotile cilia due to compound heterozygous mutations in the DNAI1 gene recovered ciliary beating after treatment with a lentivirus containing a normal DNAI1 gene . This is the first report on gene therapy in Primary Ciliary Dyskinesia . Since lentivirus is able to insert therapeutic genes into the cell genome , this result may have impact on in vivo gene therapy in this disease and in diseases related to human epithelial airway cells such as cystic fibrosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"Methods"
] | [
"respiratory",
"medicine/respiratory",
"infections",
"genetics",
"and",
"genomics/disease",
"models",
"genetics",
"and",
"genomics/gene",
"therapy",
"respiratory",
"medicine",
"genetics",
"and",
"genomics/medical",
"genetics"
] | 2009 | Ciliary Beating Recovery in Deficient Human Airway Epithelial Cells after Lentivirus Ex Vivo Gene Therapy |
Snakebites have been recognized as a neglected public health problem in several tropical and subtropical countries . Bothrops snakebites frequently complicate with acute kidney injury ( AKI ) with relevant morbidity and mortality . To date , the only treatment available for Bothrops envenomation is the intravenous administration of antivenom despite its several limitations . Therefore , the study of novel therapies in Bothrops envenomation is compelling . The aim of this study was to evaluate the protective effect of Allopurinol ( Allo ) in an experimental model of Bothrops jararaca venom ( BJ ) -associated AKI . Five groups of Wistar rats were studied: Sham , Allo , BJ , BJ+Allo , BJ+ipAllo . BJ ( 0 . 25 mg/kg ) was intravenously injected during 40’ . Saline at same dose and infusion rate was administered to Sham and Allo groups . Allo and BJ+Allo groups received Allo ( 300 mg/L ) in the drinking water 7 days prior to Saline or BJ infusion respectively . BJ+ipAllo rats received intraperitoneal Allo ( 25 mg/Kg ) 40’ after BJ infusion . BJ rats showed markedly reduced glomerular filtration rate ( GFR , inulin clearance ) associated with intense renal vasoconstriction , hemolysis , hemoglobinuria , reduced glutathione and increased systemic and renal markers of nitro-oxidative stress ( Nitrotyrosine ) . Allo ameliorated GFR , renal blood flow ( RBF ) , renal vascular resistance and arterial lactate levels . In addition , Allo was associated with increased serum glutathione as well as reduced levels of plasma and renal Nitrotyrosine . Our data show that Allo attenuated BJ-associated AKI , reduced oxidative stress , improved renal hemodynamics and organ perfusion . It might represent a novel adjuvant approach for Bothrops envenomation , a new use for an old and widely available drug .
Snakebites have been recognized as a neglected public health problem in several tropical and subtropical countries [1] . The worldwide incidence of snake envenomation may vary from 421 , 000 to 1 , 841 , 000 cases each year , resulting in up to 94 , 000 deaths [2] . However , the burden of snakebite envenomation might be underestimated , since many incidence data are derived from hospital admissions and most of the affected individuals do not seek hospital treatment [3] . Despite the nearly universal distribution , venomous snakebites are mainly found among populations living in poverty as it is inversely related to the health care expenditure [4] . These accidents predominantly affect young adults , especially males , yet almost 1/3 of the cases occur in children [5] . Moreover , snake envenomation may also be seen as an occupational hazard whereas it has been associated with significant morbidity and disability mostly in agricultural workers ( food producers ) [4–6] . Thus , there is a great impact of venomous snakebites in the economy of a community or even a country . Snakes of the genus Bothrops are the major cause of snakebites in Latin America [5] . Over 60 species distributed throughout Central and South America and the Caribbean Basin constitute the genus Bothrops [7 , 8] . In Latin America , Bothrops asper , Bothrops atrox and Bothrops jararaca are the main medically important Bothrops species [5 , 9] . Previous studies have reported substantial differences in venom composition among the genus Bothrops which might be explained by variations within species , geographic regions , dietary availability , or even by age difference considering the same animal [8 , 10–16] . However , some venom biological properties are common to all Bothrops species [8] . Bothrops venoms possess many active enzymes responsible for the local and systemic manifestations [7] . Local edema is a common finding which may be associated with drastic hemorrhagic and necrotic changes [17] . Furthermore , Bothrops venoms exhibit thrombin-like enzyme activity which activates the coagulation cascade and ultimately causes fibrinogen consumption , bleeding and disseminate intravascular coagulation [5 , 8 , 18] . Bothrops envenomation frequently complicates with acute kidney injury ( AKI ) increasing morbidity and mortality [10 , 19] . To date , the only treatment available for Bothrops envenomation is the intravenous administration of antivenom ( animal hyperimmune immunoglobulin ) . However , there are several limitations of the antivenom therapy such as high cost , limited availability , ( e . g . in remote areas ) , severe anaphylactic reactions , and inefficiency to treat local manifestations [6 , 20–22] . Therefore , the study of novel therapies in Bothrops envenomation is a fruitful area for future research . Allopurinol ( Allo ) is a xanthine oxidase ( XO ) inhibitor commonly used as uric acid ( UA ) -lowering agent . Data from several studies suggest that beneficial effects of Allo as a free radical scavenger may transcend the XO inhibition such as reducing lipid peroxidation and heat shock protein expression [23 , 24] . Additionally , treatment with Allo following rhabdomyolysis-associated AKI showed renal and muscular protective effects by reducing oxidative stress and UA levels [23] . Thus , new therapeutic applications have been accredited to Allo in the last decade particularly due to its potent antioxidant property [23 , 25–27] . Strapazzon et al . showed marked and persistent increase of several oxidative stress markers in patients hospitalized for Bothrops envenomation [28] . Therefore , the aim of this study was to evaluate the protective effect of Allo , a low-cost and widely available medication , in an experimental model of Bothrops jararaca venom ( BJ ) -associated AKI .
All experimental procedures were approved by the local research ethics committee ( “Comissão de Ética no Uso de Animais”–CEUA , 070/17 ) and developed in strict conformity with local institutional guidelines and with well-established national standards for the manipulation and care of laboratory animals ( “Resoluções Normativas do Conselho Nacional de Controle de Experimentação Animal”–CONCEA ) . Animals were anesthetized intraperitoneally with sodium thiopental ( 50 mg/kg BW ) for all the surgical experiments . After the experimental period , animals were euthanized with excess anesthetic sodium thiopental . A pool of lyophilized venom from specimens of adult B . jararaca snakes was a generous gift from the Institute Butantan , São Paulo , Brazil ( lot 01/08–01 ) . B . jararaca snakes are kept in plastic cages in temperature and humidity-controlled rooms , according the Institute Butantan Animal Care and Use Committee ( protocol n° 1296/16 ) . The venom is extracted monthly and after centrifugation the supernatant is frozen . Subsequently venoms are pooled , lyophilized , and stored at −20°C until used . On the experiment day , aliquots were dissolved in 10 ml of 0 . 9% saline for immediate use . In the present study , we aimed to use doses of Allo well established in other experimental rat studies [29–31] . Nevertheless , in view of the U . S . Food and Drug Administration ( FDA ) approach to make a dose extrapolation from humans to rats , the dose of a rat would be approximately seven times higher than the human dose ( per Kg of BW ) [32] . Thus , taking into consideration a daily dose of Allo ( 300 mg ) in an adult human weighting 60 kg ( 5 mg/Kg/day ) , the dose of Allo to be administered to the rat would be approximately 35 mg/kg/day . Considering the mean daily water intake of 25 mL/rat/day and mean BW of 272 g , we used a dose of approximately 28 mg/kg/day in the BJ+Allo group . BJ+ipAllo animals received 25 mg/Kg BW of Allo , which was around 5 times the human dose . Male Wistar rats with a mean BW of 272±5 g were obtained from the animal facilities of the University of São Paulo School of Medicine , housed in standard cages , on a 12h light/dark schedules , and given ad libitum access to water and standard diet ( Nuvilab CR-1 , Curitiba-PR , Brazil ) . Room temperature was maintained at 23°C . We administered intravenous BJ after the surgical procedure for the inulin clearance study , as previously reported by Martines et al . [6] . Animals were allocated into four groups: In order to study the therapeutic effect of Allo on renal function , we performed an additional group of animals ( BJ+ipAllo , n = 5 ) that received intraperitoneal Allo ( Sigma-Aldrich , St . Louis , USA ) ( 25 mg/kg BW ) immediately after BJ infusion . Fig 1 illustrates the experimental protocol in the first series of experiments . In separate series of experiments , using the same experimental model described above , we conducted hemodynamic studies after intravenous BJ or saline infusion ( five groups of six rats/group ) . Animals were anesthetized intraperitoneally with sodium thiopental ( 50 mg/kg BW ) . The trachea was cannulated with a PE-240 catheter , and spontaneous breathing was maintained . To evaluate the mean arterial pressure ( MAP ) and allow blood sampling , a PE-60 catheter was inserted into the right carotid artery . MAP was measured 30 minutes after the surgical procedure , using Biopac Systems Inc MP100 ( Santa Barbara , CA , USA ) . Then , a midline incision was made to measure the renal blood flow ( RBF ) . We carefully dissected the left renal pedicle and isolated the renal artery , taking precautions to avoid disturbing the renal nerves . An ultrasonic flow probe was placed around the exposed renal artery . RBF was measured using an ultrasonic flow meter ( T402; Transonic Systems , Bethesda , MD , USA ) and is expressed as ml/min . Renal vascular resistance ( RVR ) was calculated by dividing the blood pressure by RBF and is expressed as mmHg/ml/min . To determine the glomerular filtration rate ( GFR ) , we conducted inulin clearance studies . On the day of the experiment , animals were anesthetized and cannulated as above . For infusion of inulin and fluids , another PE-60 catheter was inserted into the left jugular vein . For BJ or 0 . 9% saline infusion , a PE-60 catheter was inserted into the right jugular vein . In order to collect urine samples , a suprapubic incision was made , and the urinary bladder was cannulated with a PE-240 catheter . After surgical procedure , a loading dose of inulin ( 100 mg/kg BW diluted in 0 . 9% saline ) was administered through the jugular vein . Subsequently , a constant inulin infusion ( 10 mg/kg BW in 0 . 9% saline ) was started and was continued at 0 . 04 ml/minute throughout the experiment . Blood samples and three urine samples were obtained at the beginning and at the end of the experiment . Blood and urine inulin were determined using the anthrone method . GFR data are expressed as ml/min/100g BW . At the end of either inulin clearance , organs were perfused with PBS solution ( 0 . 15 M NaCl and 0 . 01 M sodium phosphate buffer , pH 7 . 4 ) . The right kidney was removed and frozen in liquid nitrogen and stored at −80°C . The left kidney was cleaned of connective tissue and fixed in 10% formalin . Kidney sections were homogenized in an ice-cold isolation solution ( 200 mmol/l mannitol , 80 mmol/l HEPES , 41 mmol/l KOH , pH 7 . 5 ) containing a protease inhibitor cocktail ( Sigma , St . Louis , MO , USA ) , using a Teflon-pestle glass homogenizer ( Schmidt and Co . , Frankfurt am Main , Germany ) . The homogenates were centrifuged at low speed ( 4 , 000 rpm ) for 30 min at 4°C to remove nuclei and cell debris . Pellets were suspended in isolation solution with protease inhibitors . Protein concentrations were determined using the Bradford assay method ( Bio-Rad Protein Assay Kit; Bio-Rad Laboratories , Hercules , CA , USA ) . Plasma and renal nitrotyrosine ( NT ) levels , which are stable end products of peroxynitrite oxidation , were measured using a commercial ELISA kit ( HK: 501–02; Hycult Biotech , Uden , the Netherlands ) . Reduced GSH , the major endogenous antioxidant in cells , was determined in total blood by the method of Sedlak and Lindsay [33] . Whole blood was processed by addition of four volumes of ice-cold 5% metaphosphoric acid and centrifuged at 4 , 000 rpm for 10 min at 4°C . This assay consists of the reaction of supernatants of total blood samples with Ellman’s reagent to produce a yellow pigment measured spectrophotometrically at 412 nm . Serum GSH was quantified by means of the standard curve and reported as μmol/mL . Plasma sodium and potassium were measured by flame photometry ( CELM , model FC280 , São Paulo , SP , Brazil ) . Hematocrit , serum lactate and bicarbonate were determined with specific electrodes ( ABL800Flex–Radiometer , Brønshøj , Denmark ) . Fibrinogen was assessed using the Clauss modified method ( hemostasis coagulation analyzer , Stago Start 4 , France ) . Lactate dehydrogenase ( LDH ) was determined with a kinetic method using UV absorbency , which measures the conversion from L-lactate in pyruvate ( automatized analyzer , Cobas C111 , Roche , Switzerland ) . The enzymatic colorimetric method ( Labtest , Lagoa Santa , Brazil ) was used to measure plasma creatine phosphokinase ( PCPK ) and plasma uric acid ( UA ) . Hemoglobinuria was assessed using a dipstick urine test ( Cobas , Roche , Switzerland ) . Four-micrometer histological sections of kidney tissue were stained with hematoxylin–eosin ( HE ) or Masson’s trichome and examined under light microscopy . For histomorphometry , the images obtained by microscopy were captured on video via an image analyzer ( Axiovision; Carl Zeiss , Eching , Germany ) . We analyzed 30 grid fields ( 0 . 087 mm2 each ) per kidney cortex . The interstitial areas were demarcated manually , and the proportion of the field they occupied , excluding the glomeruli , was determined . In 40–60 grid fields ( 0 . 245 mm2 each; magnification , x400 ) , we graded the proportional renal damage ( tubular epithelial swelling , vacuolar degeneration , necrosis , and desquamation ) : 0 , < 5%; I , 5–25%; II , 26–50%; III , 51–75%; and IV , > 75% . To minimize bias in the morphometric analysis , the observer was blinded to the treatment groups . The mean scores were calculated by rat and by group . We used monoclonal antibody to 8-isoprostane-PGF2a ( F2-IsoP ) ( 1:500 , overnight at 4°C ) ( Oxford Biomedical Research , Oxford , England ) . We subjected 4-μm kidney tissue sections to immunohistochemical reaction according to the protocol for the primary antibody . Reaction products were detected by an avidin-biotin-peroxidase complex ( Vector Laboratories , Burlingame , CA ) , and the color reaction was developed with 3 , 3-diaminobenzidine ( Sigma Chemical ) in the presence of hydrogen peroxide . The sections were counterstained with Harris’ hematoxylin . To evaluate immunoreactivity to F2-IsoP , we analyzed 15 randomly fields of the renal medulla ( 0 . 087 mm2 each ) . The volume ratios of positive areas of renal medula ( % ) , determined by the color limit , were obtained by image analysis with the program Image-Pro Plus , version 4 . 1 ( Media Cybernetics , Silver Spring , MD ) on a computer coupled to a microscope ( Axioskop 40; Carl Zeiss ) and a digital camera . Results were expressed as percentages . Data are shown as mean ± SEM . Differences among the means of multiple parameters were analyzed by one-way analysis of variance followed by the Student–Newman–Keuls test . Values of P<0 . 05 were considered statistically significant . Data were analyzed using GraphPad Prism software 5 . 0 .
To evaluate the efficacy of prophylactic Allo on BJ-associated AKI , animals received Allo seven days prior to intravenous venom infusion . Renal function , assessed by inulin clearance , was markedly reduced in the BJ group compared to controls . In contrast , Allo pretreated rats exhibited significantly better renal function ( Fig 2A ) . Histopathological examination of kidney tissues showed glomerular microthrombi deposit in the groups that received BJ ( Fig 3A ) . Sham and Allo groups did not show any significant histopathological changes . Tubular injury score was not different between groups ( Fig 3B ) . To assess whether the potent antioxidant effect of Allo would play a role in renal protection in the current experimental model , we evaluated serum reduced GSH ( a major endogenous antioxidant ) , plasma and renal NT ( a marker of marker of nitro-oxidative stress ) . Animals from the BJ group exhibited lower levels of serum reduced GSH compared to control animals . Allo pretreatment improved serum reduced GSH levels ( Fig 2B ) . On the other hand , BJ rats exhibited higher levels of plasma and renal NT , which returned to control levels in the group pretreated with Allo ( Fig 2C and 2D ) . Overall , these data suggest an improvement in the systemic and renal redox balance conferred by Allo following B . jararaca envenomation . Renal ischemia has been considered the hallmark of AKI following B . jararaca envenomation [6 , 34 , 35] . Accordingly , we observed severe renal vasoconstriction , associated with increased renal vascular resistance , in BJ animals at 40 and 70 minutes after venom infusion . Allo pretreatment significantly improved RBF and RVR at both time points evaluated . MAP was lower in the BJ group after BJ infusion , albeit not statistically significant . Fig 4A shows a summary of the hemodynamic data . In addition , we found increased expression of F2-IsoP , a potent vasoconstrictor formed during lipid peroxidation , in the renal medulla ( Fig 4B ) . BJ animals showed significant increased expression of F2-IsoP in the renal medulla compared to other groups . Allo pretreatment reduced F2-IsoP medullary renal expression ( Fig 4B ) . Plasma sodium was not different between groups , whereas animals from BJ group showed higher plasma potassium compared to other groups ( Fig 5A and 5B ) . Furthermore , BJ infusion led to a marked decrease in fibrinogen and bicarbonate levels whilst increased serum LDH and lactate levels . BJ+Allo rats exhibited lower arterial lactate , yet fibrinogen , LDH and bicarbonate changes were not prevented by Allo pretreatment ( Fig 5C–5F ) . Moreover , Allo-pretreated rats showed lower plasma UA concentration ( Fig 5G ) . As expected , PCPK was slightly increased , albeit not statistically significant , after intravenous BJ administration ( Fig 5H ) . Although we found hemoglobinuria in animals that received BJ , there were no differences on hematocrit levels between groups ( Fig 5I and 5J ) . We have conducted additional studies using intraperitoneal Allo after BJ infusion in order to assess if therapeutic Allo also improved the functional parameters ( renal function , RBF and RVR ) . Strikingly , therapeutic Allo also improved renal function ( 0 . 74 ± 0 . 05 mL/min/100g BW; p<0 . 01 vs . BJ group ) , RBF ( 4 . 4 ± 0 . 5 mL/min; p<0 . 01 vs . BJ group ) and RVR ( 18 . 5 ± 1 . 5 mmHg/mL/min; p<0 . 001 vs . BJ group ) following BJ envenomation .
AKI seems to be a major cause of death among patients who survived the early BJ-induced systemic effects [10] . Animal models of AKI following intravenous infusion of BJ have been validated to study the mechanisms involved in this syndrome as well as potential therapeutic agents [6 , 34 , 35] . Similar to previous studies , we found that intravenous administration of BJ caused striking reduction of renal function associated with intense renal vasoconstriction , fibrinogen consumption and intravascular hemolysis without systemic blood pressure changes . We also reported that prophylactic and therapeutic Allo attenuated renal dysfunction secondary to BJ infusion . To our knowledge , this is the first demonstration of Allo as both treatment and prevention of BJ-associated AKI . In recent years , there has been an increasing amount of literature on the role of oxidative stress in the pathogenesis of snake envenomation [28 , 36 , 37] . Snake venoms present a series of biologically active peptides , such as L-amino acid oxidase ( LAAO ) and phospholipase A2 ( PLA2 ) which might be related to the generation of reactive oxygen species ( ROS ) [38 , 39] . LAAOs are flavoenzymes widely distributed among venomous snake families that catalyze the oxidative deamination of L-amino acid to α-keto acids , thus , releasing hydrogen peroxide and ammonia [39] . Venom PLA2 , in contrast to endogenous PLA2 , has been linked with a wide spectrum of toxic effects , leading to increased systemic oxidative stress [39 , 40] . Therefore , to evaluate the involvement of the oxidative stress in the pathogenesis of BJ-associated AKI , as well as its inhibition by Allo therapy , we assessed serum reduced GSH , plasma and renal NT . GSH is the most prevalent intracellular thiol , whose main role is to protect cells from oxidative damage [41] . On the other hand , NT is a byproduct of protein tyrosine residues generated from the reaction of nitric oxide ( NO ) and superoxide ( O2•— ) [42] . In recent years , it has been demonstrated the important role of NO regulating the redox signaling [43] . When reacted with O2•— , NO is a major substrate to generate reactive nitrogen species which can lead to nitration and nitrosation of several substrates , mostly proteins with various physiological functions [43] . This biological process has been called by some authors as nitro-oxidative stress [43–45] . Furthermore , Walker et al . described renal NT as an early marker of oxidative stress in the setting of renal ischemia reperfusion [46] . We found evidence of redox imbalance in the rats from BJ group . BJ infusion was associated with reduced levels of GSH levels and increased systemic and renal markers of nitro-oxidative stress , whilst Allo pretreatment improved all these markers . Renal vasoconstriction is one of the key elements in the pathogenesis of BJ-associated AKI [6 , 34 , 35] . Martines et al . showed a significant reduction of RBF and increase of RVR following BJ infusion [6] . Furthermore , previous studies have highlighted the pivotal role of isoprostanes in oxidative stress-mediated renal vasoconstriction [23 , 47 , 48] . Accordingly , we found profound renal vasoconstriction combined with increased expression of F2-IsoP in the renal medulla . We believe that two factors might have contributed to this finding . First , given the high sensitivity of the medulla to decreased O2 supply , it is reasonable to expect that it would be the first portion of the kidney affected by vasoconstriction and oxidative stress [49] . Second , since we performed an early assessment of F2-IsoP ( 90’ after BJ infusion ) , we might not have been able to detect cortical alterations at this stage . Allo ameliorated renal hemodynamic changes in this experimental model . We suggest that Allo might have inhibited the lipid peroxidation and consequently reduced F2-IsoP formation . Conversely , F2-IsoP may represent a marker of increased oxidative stress following intense vasoconstriction in the renal medulla . Nevertheless , we could not provide evidence of causality between F2-IsoP and BJ-induced renal vasoconstriction and this is still an important issue for future research . In addition to the renal hemodynamic changes , fibrin thrombi deposits in the glomerular capillary may also play a role in the pathogenesis of BJ-associated AKI contributing to the reduction in GFR and RBF [34] . Hence , data from necropsy kidney samples showed glomerular fibrin thrombi following Bothrops snakebites in individuals with cortical necrosis [50] . We showed glomerular microthrombi deposit in the groups that received BJ . Nevertheless , tubular injury score was not different between groups , which might be due to the renal tissue harvesting in the very early stage of AKI , therefore , too early to observe significant renal histological damage . On the other hand , we did not find significant difference in CPK levels among the study groups , which speaks against rhabdomyolysis as a pathogenic factor for AKI in this experimental model . In fact , Burdmann et al . also reported no increase in CPK levels using similar animal model . We believe that rhabdomyolysis is mainly a local manifestation of BJ envenomation , thus , occurring after subcutaneous or intramuscular inoculation of the venom rather than with intravenous injection . In the past years , numerous studies have reported the pro-oxidant properties of UA [51–53] . Higher UA levels may activate NADPH oxidase , increasing protein nitrosation and lipid peroxidation [54] . Moreover , several lines of evidence suggest an association between UA with hypertension , cardiovascular disease , progression of chronic kidney disease and rhabdomyolysis-associated AKI [23 , 55–57] . In the present study , we assume that the positive effects of Allo resulted predominantly from free radical scavenging . However , we may not neglect the potential adjuvant effect of UA-lowering therapy . Therefore , further research is warranted to uncover the role of UA in the pathogenesis of BJ-associated AKI . Bothrops venoms are composed of several different substances related to broad range of systemic effects [7] . Intravenous infusion of BJ has been linked with fibrinogen consumption , intravascular hemolysis and shock ( reduced organ perfusion ) . Similarly , our animal model reproduced all the systemic effects of BJ envenomation as described in previous studies [6 , 34 , 35] . However , one limitation of our experimental model was the short follow up after BJ infusion . This might be the reason why hemoglobin levels as well as MAP were not significantly lower in the BJ group compared to other groups . In fact , we believe that we might have observed a period of cryptic shock ( hyperlactatemia without persistent hypotension ) associated with initial features of hemolysis ( increased LDH and hemoglobinuria ) and hemoconcentration , the latter also described in early stages of hemorrhagic shock [58] . Overall , BJ+Allo rats showed better organ perfusion ( i . e . lower arterial lactate levels ) which is in accordance with previous studies that reported protective effects of Allo in experimental models of hemorrhagic shock by improving hemodynamics , preserving tissue levels of ATP and ultimately reducing mortality [59 , 60] . Therefore , the studies discussed here highlighted the role of Allo beyond free radical scavenging in the setting of shock . In conclusion , our study show that Allo attenuated BJ-associated AKI , reduced oxidative stress , improved renal hemodynamics and organ perfusion . We hope that this study encourages further clinical investigation regarding the beneficial effects of Allo on Bothrops envenomation . If supported by clinical studies , it might represent a novel adjuvant approach for Bothrops envenoming , a new use for an old and widely available drug . | Snakebites have been recognized as a neglected public health problem in several tropical and subtropical countries . Snakes from the genus Bothrops are the most common snakes found in Latin America . Bothrops snakebites frequently complicate with kidney impairment with increased risk of death . The only treatment currently available for Bothrops envenomation is the infusion of antivenom . Therefore , the study of novel therapies to prevent kidney failure caused by snakebites is compelling . Allopurinol is a drug commonly used to treat gout which also presents beneficial antioxidant properties . The aim of this study was to evaluate the protective effect of Allopurinol in a rat experimental model of Bothrops envenomation . In our study , rats that received the infusion of the snake venom presented with reduced kidney function which was improved by Allopurinol . Overall , Allopurinol might represent a novel approach to prevent kidney failure in victims of Bothrops snakebites . This could be a new use for an old and widely available drug . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"toxins",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"oxidative",
"stress",
"tropical",
"diseases",
"vertebrates",
"animals",
"toxic",
"agents",
"toxicology",
"reptiles",
"vasoconstriction",
"neglected",
"tro... | 2017 | Allopurinol attenuates acute kidney injury following Bothrops jararaca envenomation |
The construction of a large dendritic arbor requires robust growth and the precise delivery of membrane and protein cargoes to specific subcellular regions of the developing dendrite . How the microtubule-based vesicular trafficking and sorting systems are regulated to distribute these dendritic development factors throughout the dendrite is not well understood . Here we identify the small GTPase RAB-10 and the exocyst complex as critical regulators of dendrite morphogenesis and patterning in the C . elegans sensory neuron PVD . In rab-10 mutants , PVD dendritic branches are reduced in the posterior region of the cell but are excessive in the distal anterior region of the cell . We also demonstrate that the dendritic branch distribution within PVD depends on the balance between the molecular motors kinesin-1/UNC-116 and dynein , and we propose that RAB-10 regulates dendrite morphology by balancing the activity of these motors to appropriately distribute branching factors , including the transmembrane receptor DMA-1 .
There is great diversity in the structure and complexity of dendritic arbors across neuron types , and establishing the correct dendritic morphology is critical for the proper connectivity and function of neural circuits . A developing dendritic arbor must target a specific receptive field , adopt the appropriate neuron-specific architecture , and avoid overlapping in connectivity with itself and neighboring dendrites . A number of extrinsic cues and intrinsic mechanisms help orchestrate the formation of these complex neuronal morphologies , including transcriptional programs , extracellular guidance cues , and contact-dependent repulsive molecules that mediate self-avoidance [1–5] . Dendritic arbor development requires tremendous cellular growth and likely has specialized membrane trafficking demands . Little is known about how the transport of branching factors and membrane components is coordinated across a large , polarized neuron . Dendrites are more sensitive than axons in their reliance on the membrane supply from secretory pathways [2 , 6] , and they have distinct transport needs . For example , a set of dendritic arbor reduction ( dar ) genes have been identified which are required for dendritic arbor outgrowth but not axon growth . These dar genes are important for ER-to-Golgi transport [6] . In addition , the Rab GTPases , a conserved family of small GTPase proteins that regulate membrane identity and vesicle trafficking [7–9] , are likely important for the polarization and outgrowth of neurites , though their precise role in both axonal and dendritic development remains unclear [10] . One of these small GTPases , Rab10 , has been shown to mediate membrane trafficking in several polarized cell types , including neurons [11–18] . The importance of Rab10 for endosomal sorting and endocytic recycling has been demonstrated in Drosophila epithelial cells [15] as well as in C . elegans neurons [16–17] and intestinal epithelial cells [11–14] . In hippocampal neurons , Rab10 is required for directional membrane trafficking in the growing axon [19–20] , and it has been shown to directly bind the anterograde motor kinesin1 via the adaptor protein JIP1 [21] . In addition to sorting and trafficking cargo to the appropriate destinations , a growing neurite must appropriately dock and release membrane and protein cargoes . The docking of Rab10-postive vesicles is required for axon outgrowth in hippocampal neurons [22] . Rab10 has also been linked to the exocyst , a secretory complex responsible for polarized exocytosis [23–25] . The exocyst is an octameric complex that was initially identified in yeast for its role in membrane addition during bud outgrowth , and it has since been shown to have conserved functions in polarized growth across several cell types [25–28] . In Drosophila , the sec5 component of the exocyst complex is required to establish polarity in the oocyte [26] , and the sec15 component is also required for the polarized delivery of photoreceptors [28] . Both Rab10 and the exocyst complex are required for branch outgrowth in the Drosophila trachea [25] . Taken together , these data suggest that both the exocyst complex and RAB-10 support outgrowth in polarized cells; however , the roles of RAB-10 and the exocyst have never been examined in a growing dendrite . In Caenorhabditis elegans , the PVD sensory neuron forms elaborate dendritic arbors that are organized into orthogonal tiers of characteristic menorah-shaped units [29–30] . The PVD neuron has proved to be useful in the study of the molecular mechanisms of dendrite morphogenesis . Importantly , the dendritic branches of PVD maintain self-avoidance [31] , a key feature across neural networks . Extracellular cues are essential for PVD dendritic morphogenesis: the branching receptor DMA-1 , expressed in PVD , instructs spatially-restricted growth of dendritic branches via interaction with the SAX-7/MNR-1 adhesion complex [32–34] . Previous work has also identified various intracellular processes important for PVD dendritic morphology , including transcription factors that instruct cell fate [29 , 35] , microtubule-associated proteins [36] , and membrane fusion proteins that prevent over-branching [37] . However , the role of polarized membrane trafficking in the growing PVD dendrite remains uncharacterized . In this work , we identified a novel role for the small GTPase RAB-10 in patterning the dendritic arbor of PVD . We demonstrate that both RAB-10 and the exocyst complex are required for proper outgrowth and patterning of the PVD dendritic arbor . RAB-10 and EXOC-8 are also required for the appropriate localization of branching factors such as DMA-1 . Additionally , we found that the molecular motors kinesin and dynein are important for anterior-posterior patterning of the PVD dendritic arbor . The PVD dendritic outgrowth phenotype was also observed in a recently published report [38] , and our work further identifies a role for RAB-10 in dendritic patterning beyond its role in promoting outgrowth . We propose that RAB-10 regulates dendrite morphogenesis by balancing anterograde and retrograde transport via molecular motors .
We visualized the morphology of the PVD neurons by expressing membrane-bound-GFP under the control of a cell-specific promoter ( ser2prom3:myrGFP ) ( Fig 1A and 1B ) [29] . In wild-type animals , one anterior and one posterior primary dendritic branch extend from the cell body during the L2 larval stage . Dense menorahs consisting of a series of orthogonal 2° , 3° , and 4° branches emerge from the anterior and posterior 1° dendrites as the animal progresses through subsequent larval stages ( L2-L4 ) ( Fig 1A , 1B and 1E ) . To understand the molecular mechanisms of dendritic branching , we performed a forward genetic screen to identify mutations that affect the dendritic morphology of PVD . From this screen , we isolated individuals with the fully penetrant and recessive mutant allele wy787 , which show a severe reduction in the number of dendritic branches ( Fig 1C and 1D ) . wy787 mutant animals also have several non-neuronal phenotypes , including intestinal vacuoles and reduced brood size . In wy787 mutants , the number of secondary branches is reduced from 42 ± 6 in wild-type to 28 ± 9 in wy787 , and the number of quaternary branches is reduced from 114 ± 15 to 17 ± 10 ( S1 Table ) . In addition to the reduction in branch number , we noticed a shift in branch distribution . In wy787 mutants , considerably more branches are found in the distal anterior region of the 1° dendrite , where wild-type animals typically grow few branches ( Fig 1B and 1D ) . To quantify the distribution of menorah , we divided the entire PVD dendritic arbor into four regions: the region posterior to the cell body ( designated as “-1” ) and three equal-length regions in the anterior dendrite ( designated as “+1 , +2 and +3” ) . We defined a “branch complexity” index to characterize the completeness of menorahs in each region . This branch complexity metric weights three characteristics equally: ( 1 ) the number of secondary branches , ( 2 ) the percentage of secondary branches that form a tertiary branch , ( 3 ) and the average number of quaternary branches per tertiary branch ( Fig 1G ) . In wy787 mutants , branch complexity is severely reduced in the +2 , +1 , and -1 regions but increased in the +3 region . ( Fig 1C , 1D and 1H ) . The severe decrease in branch complexity in the +1 and -1 regions is the result of a near-total loss of 3° and 4° branches in the +1 and -1 regions ( S1 Table ) . The increase in branch complexity in the anterior +3 region led us to hypothesize that the branching defects cannot be entirely explained by an overall lack of branching activity , but are instead the result of abnormal distribution of branching activity along the anterior-posterior axis . The lack of branches in the posterior regions ( +1 and -1 ) could be due to a defect in 2° branch outgrowth or stabilization . We distinguished between these possibilities by examining dendrite morphology during earlier developmental time points . In the L2 and L3 larval stages , wild-type and wy787 showed no significant difference in the number of secondary filopodia ( Fig 1F ) , which suggests that wy787 could be defective in branch-stabilization . However , in wy787 the posterior 2° branches are often shorter than they are in wild-type and often do not grow in the correct orthogonal orientation . These defective branches cannot reach the tertiary branch-point ( Fig 1C ) , a region enriched for the branch-promoting complex , MNR-1/SAX-7 , that supports tertiary branch outgrowth [33] . By the L4 stage , wy787 had significantly fewer 2° branches , and those 2° branches that reached the tertiary line in the posterior regions ( +1 and -1 ) failed to form 3° branches ( Fig 1D and 1F ) . This developmental failure is consistent with a defect in both stabilization and outgrowth of 2° branches as well as both initiation and maintenance defects in 3° and 4° branches . This defective branching in the posterior regions is reminiscent of the phenotype of the branching receptor mutant dma-1 [32–33] . We mapped and cloned the wy787 allele and identified the causal lesion as a mutation in the small GTPase RAB-10 . wy787 contains a point mutation that results in a substitution of a conserved residue ( A66V ) in the GTP-binding domain of RAB-10 ( S1 Fig ) and can be rescued by rab-10 genomic DNA ( S1 Fig ) and rab-10 cDNA ( Fig 2B ) . To confirm that RAB-10 is required for dendritic arbor growth , we also obtained the putative null allele ok1494 , which has a 663 base pair deletion in the rab-10 coding region and should cause a complete loss of function of this gene ( Fig 2E ) . This deletion allele recapitulated the PVD dendritic phenotype of wy787 with an even more severe reduction in posterior branch complexity and a more anterior shift in the +3 region compared with the wy787 allele ( Fig 2C ) . To determine whether RAB-10 functions cell-autonomously within PVD to pattern the dendritic arbor , we expressed RAB-10 under the control of two PVD-specific promoters ( ser2prom3 and pdes2short ) [30] in the wy787 and ok1494 mutant backgrounds . Expression of RAB-10 in PVD rescued the posterior reduction in branch complexity ( Fig 2B , 2D , 2F and 2G ) , while muscle-specific expression of RAB-10 failed to rescue the dendritic phenotype of rab-10 mutants ( S1 Fig ) . Taken together , these data suggest that RAB-10 functions cell-autonomously to regulate the dendritic branch distribution of PVD neurons . RAB-10 is a member of the Rab family of GTP-binding proteins . These conserved proteins help to control membrane identity and regulate intracellular vesicle trafficking [7–9] . A typical Rab GTPase adopts two conformations: an active GTP-bound form and an inactive GDP-bound form . Guanine nucleotide exchange factors ( GEFs ) help activate Rabs by catalyzing the exchange of GDP for GTP , and GTPase activating proteins ( GAPs ) accelerate the GTPase activity of Rabs [7] . The missense mutation in wy787 , A66V , is two amino acids away from the GTP binding site of RAB-10 and likely interferes with the innate activity of RAB-10 . GTP-binding sites and GTPase sites are highly conserved across Rab GTPases in many species ( S1 Fig ) , and it has been well established that these sites can be mutated to generate dominant-negative and constitutively active forms of Rab GTPases [11–14 , 17–18 , 20 , 23] . To assess whether the GTPase activity of RAB-10 is required for normal PVD development , we generated the constitutively active and inactive forms of RAB-10 . The constitutively active mutant , Q68L , alters the GTPase site to prevent GTP-hydrolysis , thus locking RAB-10 in its active conformation . The inactive form , T23N , prevents GTP-binding and should result in an inactive RAB-10 . Using a PVD-specific promoter , we expressed either the active or inactive form of RAB-10 in both wild-type and the rab-10 ( ok1494 ) mutant background . In the wild-type background , the putative GDP-locked ( T23N ) of RAB-10 likely acts as a dominant negative . We observed a moderate reduction in posterior branch complexity in 30% of animals , and 1 in 50 animals have a branching phenotype indistinguishable from that of rab-10 ( S1 Fig ) . Interestingly , overexpression of rab-10 ( GDP ) also causes an anterior shift in branch complexity . This provides further support for the cell-autonomous action of RAB-10 . The putative GTP-locked form ( Q68L ) of RAB-10 rescued the dendritic morphology defect of rab-10 ( Fig 3A and 3C ) , though the rescue was not as robust as that of the wild-type form of RAB-10 . The dominant-negative RAB-10 ( GDP ) not only failed to rescue the dendritic defect but also further reduced branch complexity in the anterior region ( Fig 3B and 3D ) . This is likely a dominant negative effect caused by sequestration of GEFs that interact with RAB-10 and one or more additional small G proteins . Together , these results support the notion that RAB-10 functions cell-autonomously as a GTPase to regulate PVD dendrite development . The DennD4 protein has been shown to act as the GEF for the mammalian RAB10 protein [39] . We asked whether the worm DennD4 homolog , denn-4 , is also required for PVD morphogenesis . Since the existing denn-4 alleles result in early lethality , preventing us from studying PVD morphology , we used somatic CRISPR-Cas9 to generate a conditional denn-4 knockout ( Fig 4A ) . We used a heat shock promoter ( Phsp16 . 2 ) to drive the expression of CAS9 and an sgRNA designed to target denn-4 , a method that has been proven effective in generating somatic mosaics [40] . Phsp16 . 2 is active in many tissues , including neurons [41] . Indeed , knocking down denn-4 with this strategy caused a reduction of branch complexity in the -1 and +1 regions in 21% of animals ( Fig 4C–4E ) . While this phenotype is milder than the branch reduction seen in rab-10 , it is qualitatively similar , suggesting that denn-4 acts as a GEF for RAB-10 and to promote dendrite morphogenesis . In order to establish the relationship between denn-4 and rab-10 , we knocked down denn-4 using somatic CRISPR-Cas9 while simultaneously overexpressing RAB-10 ( GTP ) in PVD . Overexpression of RAB-10 ( GTP ) reduced the penetrance of denn-4 dendritic phenotypes to 6% ( Fig 4D ) . Since Phsp16 . 2 likely knocks down denn-4 in many tissues , we cannot determine whether denn-4 functions cell-autonomously within PVD . However , suppression of the denn-4 phenotype by PVD-specific expression of RAB-10 ( GTP ) supports the notion that denn-4 plays a role in dendritic development within PVD , and that denn-4 acts as a GEF of rab-10 , as suggested by studies in non-neuronal cells [39] . Together , these data argue strongly that the GTPase activity of RAB-10 is required to regulate dendrite morphogenesis in PVD . RAB-10 has been shown to function with the exocyst complex in vesicle trafficking and docking [23–25] . Therefore , we tested whether rab-10 acts via the exocyst to regulate dendrite branch distribution in PVD . We examined PVD morphology in strains carrying mutations in each of several conserved exocyst genes , including exoc-8 and sec-5 . The exoc-8 ( ok2523 ) allele is a 1474 base pair deletion and a putative null [24] . The sec-5 ( pk2358 ) allele introduces an early stop codon at position 389 , and has been described as a non-null , strong loss-of-function allele [42] . We found that mutations in either exoc-8 ( ok2523 ) or sec-5 ( pk2358 ) resulted in a posterior loss of dendritic branch complexity , nearly phenocopying rab-10 mutants ( Fig 5A and S2 Fig ) . The reduction in secondary branches in the +2 region was less dramatic in the exoc-8 ( ok2523 ) mutant than in the rab-10 mutants ( S1 Table ) , but significant reduction in branch complexity was observed in the +2 , +1 , and -1 regions ( Fig 5A and 5F ) . sec-5 ( pk2358 ) displayed a less severe but qualitatively similar phenotype ( S2 Fig ) . No significant 2° branch loss was observed in sec-5 ( S1 Table ) , but 4° branches were reduced in the +2 , +1 , and -1 regions ( S1 Table ) . To test whether the exocyst complex functions cell-autonomously to pattern the PVD dendrite , we expressed exoc-8 using a PVD specific promoter in the exoc-8 ( ok2523 ) background . Expressing exoc-8 in PVD rescued the dendritic morphology defect ( Fig 5B and 5F ) , suggesting that exoc-8 also functions cell-autonomously . The striking anterior shift of dendritic complexity observed in rab-10 was not observed in exoc-8 ( ok2523 ) or sec-5 ( pk2358 ) mutants ( Fig 5A and S2 Fig ) . While both rab-10 and the exocyst are required for branching in the normal +2 , +1 , and -1 regions , this indicates that rab-10’s role in anterior-posterior patterning may not require the exocyst components . We investigated the relationship between rab-10 and the exocyst with two genetic interaction experiments . Since rab-10;exoc-8 animals do not survive [24] , we overexpressed the dominant-negative RAB-10 ( GDP ) in exoc-8 to assess the double mutant phenotype . Overexpression of RAB-10 ( GDP ) in exoc-8 further reduced branch complexity to rab-10 levels and also caused an anterior shift in branch complexity , phenocopying rab-10 ( Fig 5E ) . The exoc-8;RAB-10 ( GDP ) phenotype was indistinguishable from that of rab-10 and not stronger than the rab-10 phenotype ( Fig 5G ) , suggesting that rab-10 and the exocyst function in a common pathway . To further explore this genetic relationship , we overexpressed RAB-10 ( GTP ) in the exoc-8 background . RAB-10 ( GTP ) was able to rescue the +2 , +1 , and -1 reduction in branch complexity caused by the exoc-8 mutation ( Fig 5C and 5F ) , suggesting that rab-10 acts downstream of exoc-8 . Together these data demonstrate that both rab-10 and the exocyst act cell-autonomously to pattern the PVD dendrite . rab-10 likely has exoc-8 independent functions in anterior-posterior patterning , but these genetic interaction data indicate that rab-10 and the exocyst function in a common pathway to promote proximal dendritic growth . The lack of stabilized dendritic branches , despite secondary filopodia outgrowth , in the rab-10 mutants is similar to phenotypes found in the dma-1 mutants [32–33] . We have previously demonstrated that DMA-1 functions cell autonomously in PVD as the branching receptor that responds to extrinsic SAX-7/MNR-1 guidance cues to promote 3° and 4° branch formation as well as stabilization of 2° branches [33] . We observed that in dma-1 mutants , all 4° and most 3° branches are lost in all regions of PVD ( S3 Fig and S1 Table ) . Since DMA-1 is a transmembrane protein and functions as a receptor on the plasma membrane , we hypothesized that rab-10 and exoc-8 might regulate the membrane localization of DMA-1 . To understand how RAB-10 and EXOC-8 regulate DMA-1 localization , we expressed a DMA-1::GFP fusion protein under the control of a PVD specific promoter in wild-type , exoc-8 and rab-10 animals . In wild-type animals , DMA-1::GFP displays a predominantly diffuse staining pattern throughout the entire PVD dendrite , which likely represents DMA-1 localization on the plasma membrane . DMA-1::GFP can also be observed as punctate structures that likely represent secretory and endocytic vesicles ( Fig 6A ) . The majority of DMA-1 puncta co-localize with the late endosomal marker RAB-7 , suggesting that they represent RAB-7-positive vesicles targeted for lysosomal degradation ( S4 Fig ) . This co-localization is not disrupted in rab-10 , but RAB-7 forms larger accumulations than the normal wild-type puncta in the primary dendrite of rab-10 ( S4 Fig ) . A small portion of the DMA-1 puncta overlap with RAB-10 puncta ( S4 Fig ) . In rab-10 , the diffuse component of DMA-1::GFP appears absent from the posterior dendrite ( +2 , +1 and -1 regions ) and the DMA-1::GFP puncta form larger accumulations than in wild-type , similar to the change seen in RAB-7 localization ( Fig 6B and S3 Fig and S4 Fig ) . In the anterior ( +3 ) region of PVD , where the full dendritic arbor forms normally , DMA-1::GFP localization appeared largely normal in the rab-10 mutant ( Fig 6B ) . Since both exoc-8 and rab-10 are required for posterior dendritic growth , we also examined DMA-1::GFP localization in exoc-8 . Similar defects in DMA-1::GFP localization were observed , with a decrease in diffusive staining in the +1 region and large accumulations of intracellular material in the primary and defective secondary branches . ( Fig 6C ) . To further examine this localization defect , we used fluorescence recovery after bleaching ( FRAP ) assays to examine the lateral mobility of DMA-1::GFP . We reasoned that the DMA-1 on the plasma membrane should be more diffusive compared with the DMA-1 trapped in intracellular membrane organelles . We focused on the primary dendrite in the +1 region because this region showed pronounced differences between wild-type and rab-10 in both DMA-1::GFP localization and dendritic arbor formation . Indeed , wild-type animals showed a more complete and faster recovery of fluorescence than in the rab-10 mutants ( Fig 6D–6F ) , further supporting the notion that DMA-1 fails to be inserted into the plasma membrane in the posterior dendrite of the rab-10 mutants . In dma-1 mutants , the PVD dendrite is almost completely devoid of 2° , 3° , and 4° branches [32–34] ( S3 Fig ) , making the branching defect more severe than that of rab-10 . This implies that DMA-1 function is not completely dependent on rab-10 . Conversely , if rab-10 controls PVD dendrite morphogenesis only by regulating DMA-1 , we would predict that the dma-1; rab-10 double mutant should phenocopy the dma-1 mutant . In the dma-1;rab-10 double mutant , 3° and 4° branches are lost in all regions of PVD , including the distal regions ( S3 Fig and S1 Table ) , and in this regard it is indistinguishable from the dma-1 single mutant . However , the double mutant has fewer 2° branches than either rab-10 or dma-1 alone ( S3 Fig and S1 Table ) . This further reduction in 2° branches suggests that rab-10 may regulate a larger cohort of proteins , which includes DMA-1 , that are important for branching activity . To assess rab-10’s general role in distributing membrane proteins , we also examined the exogenous membrane protein mcd8 . 3 in both wild-type and rab-10 animals and observed a similar phenomenon . More accumulations of large puncta , which likely represent intracellular protein , occurred in rab-10 than in wild-type , and the diffuse proportion was reduced . This defect was most notable in the incomplete secondary branches ( S3 Fig ) . The normal subcellular localization of DMA-1 and excessive branches in the anterior dendrite of the rab-10 mutant argue strongly that RAB-10 is not essential for the function of DMA-1 . Instead , RAB-10 is required for correctly distributing “branching activity” along the anterior-posterior axis of the primary dendrite . This branching activity includes both DMA-1 and other factors required for dendrite branching . Even in the wild-type animals , the branches are not evenly distributed across the entire 1° dendrite: the anterior +3 region contains far fewer branches compared with the +2 or +1 region , suggesting a regulated distribution of branching activity ( Fig 1B and 1H ) . It is therefore plausible that RAB-10 is critical to establish this distribution . Since microtubule-based transport is essential for organelle and protein localization in large cells like PVD , we characterized the microtubule organization along the primary dendrite . Using the established EBP-2::GFP marker [43] to visualize the growing tips of microtubules in different regions of PVD , we found that microtubules in the anterior primary dendrite ( +1 , +2 and +3 regions ) are predominantly oriented minus-end out with respect to the soma , with the minus end pointing to the distal anterior dendrite ( Fig 7A and 7C ) . This is consistent with the literature that suggests most invertebrate dendrites adopt minus-end out microtubule organization [44–46] . Surprisingly , microtubules in the posterior dendrites ( -1 region ) are mostly plus-end out ( Fig 7A and 7C ) , mimicking the organization of polarity in axons ( Fig 7A and 7C ) . This microtubule organization dictates that a retrograde minus-end-directed motor , such as dynein , should move processively from the posterior-most dendrite to the anterior-most dendrite . Conversely , an anterograde plus-end-directed motor , such as kinesin-1 , would move from the anterior-most dendrite to the posterior-most region . We examined the PVD phenotype in dhc-1 and unc-116/kinesin-1 mutants to directly test if major molecular motor systems are involved in PVD dendrite branching , possibly by transporting vesicles carrying DMA-1 and other dendrite outgrowth factors into the developing dendrite . Similar to previous a report , we found that the dhc-1 mutant lacked dendritic branches specifically in the anterior region [36] . This is consistent with what we would expect based on microtubule organization in PVD , because dynein transports cargoes towards the anterior dendrite . Reduced dynein activity in the dhc-1 mutant might result in the specific reduction of branching activity in the distal anterior dendrite . Strikingly , consistent with our predictions , the posterior dendritic branches showed no reduction in complexity in the dhc-1 mutant ( Fig 8B ) . Next , we examined the unc-116/kinesin-1 mutant for PVD dendrite morphology defects . Similar to previous reports , we found that a partial-loss-of-function unc-116 allele ( e2310 ) caused a severe lack of anterior branches [36] . This phenotype is especially pronounced in the distal anterior region ( Fig 8C ) . Superficially , this result is not consistent with our model; UNC-116/kinesin-1 should traffic cargoes toward the posterior part of the cell . However , our previous experiments in the neuron DA9 showed that UNC-116 is required for establishing microtubule polarity in the dendrite in addition to its known function in trafficking organelle cargoes . In the unc-116 mutant , the DA9 dendrite adopts an axon-like “plus-end distal” organization , while the axonal microtubules are not affected [43] . To test if the unc-116 mutant also alters microtubule polarity in the PVD dendrite , we visualized EBP-2::GFP comets in anterior and posterior primary dendrites . Similar to the result seen in the DA9 dendrite , the microtubules in the anterior PVD dendrites are reversed in orientation compared to the wild-type controls: the microtubules of the anterior dendrite switch from minus-end distal in wild-type to predominantly plus-end distal in the unc-116 mutant ( Fig 7B , 7C and 7D ) . Interestingly , posterior dendrites in the mutant maintain the plus-end distal orientation , similar to the wild-type control ( Fig 7B , 7C and 7D ) . As a result of the polarity change in the partial-loss-of-function allele , kinesin/UNC-116 should traffic cargoes to the distal anterior and distal posterior dendrite , while dynein should traffic cargoes toward the cell body ( Fig 7D ) . Therefore , in the unc-116 mutants , both the reversal in microtubule polarity and the reduced activity of the kinesin/UNC-116 motor might contribute to the dendritic phenotype . It is conceivable that the reduced kinesin activity , normal dynein activity , and plus-end distal microtubule orientation leads to a specific loss of branching cargo from the anterior distal dendrite . Together , these data demonstrate that microtubules in both anterior and posterior primary dendrite in wild-type animals are oriented such that the vast majority of minus-ends of microtubules point to the anterior end of the animal ( Fig 7A ) . The dramatic anterior shift of dendritic arbor observed in rab-10 suggested that RAB-10 could be responsible for patterning the dendritic arbor by directing protein trafficking along the A-P axis . It has previously been demonstrated that vertebrate Rab10 interacts with kinesin-1 to regulate axonal vesicle transport in the growing axon of hippocampal neurons [21] . Based on the anterior bias in branch distribution observed in the rab-10 mutants , we hypothesized that RAB-10 promotes the transport of branching activity towards the plus ends of microtubules . To further test this hypothesis , we explored the genetic interaction between rab-10 and unc-116 . The reversed polarity of microtubules in the anterior dendrite of unc-116 mutants provided a framework to test the potential cooperation between rab-10 and unc-116 . If the anterior shift of PVD dendritic branches seen in rab-10 occurred as a result of a failure of plus-end directed trafficking , then we would expect the branching defect to be reversed by switching microtubule polarity: our hypothesis predicts that the double mutants should lack distal anterior dendrites but have branches in the anterior +1 and posterior -1 region . Indeed , we found that the double mutant showed reduced branch complexity in the +3 region compared with that of rab-10 single mutants ( Fig 8E ) . In striking contrast to the rab-10 single mutants , which have their highest branch complexity in the anterior “+3” region , the double mutant has the highest branch complexity in the posterior “+1” and “-1” regions , where the microtubule minus-ends are directed ( Fig 8E and 8G ) . To further explore the relationship between rab-10 and the molecular motors , we examined the effect of knocking down dynein in the rab-10 background . In our model , the minus-end-directed dynein motor is predicted to be responsible for distributing branch activity to the distal anterior dendrite , and our hypothesis predicts that the loss of dynein would suppress the anterior-shifted branches of rab-10 . We utilized the previously described somatic CRISPR-Cas9 system to knock down dhc-1 . The conditional dhc-1 knockout in the wild-type background resulted in a reduction of branch complexity in the distal anterior region , similar to the or195ts allele ( Figs 8B , 9A and 9D ) . The conditional dhc-1 knockout in the rab-10 background resulted in a complete loss of branches from the distal anterior region ( Fig 9C and 9E ) in 57% of animals , consistent with our hypothesis that dynein is required for distributing distal branching activity . These genetic interaction experiments between RAB-10/UNC-116 and RAB-10/DHC-1 provide compelling support for RAB-10’s role in regulating motor-assisted transport of factors required for branching activity , and this suggests that RAB-10 is responsible for determining not just outgrowth , but also the distribution of branches . We propose that RAB-10 plays dual functions in regulating dendritic branching . On one hand , it regulates dendritic trafficking of branching factors towards the plus-ends of microtubules along the A-P axis , likely through association with UNC-116/kinesin-1 . The factors regulated by RAB-10 include the branching receptor DMA-1 and other unidentified membrane components . In addition , once the vesicles are delivered to the proper domain of the plasma membrane , RAB-10 is required to dock and fuse membrane vesicles through its interaction with the exocyst complex .
Dramatic dendritic morphogenesis defects were observed in the posterior regions of PVD in the rab-10 mutant , which were also independently described in a recent report [38] . However , our detailed analysis of PVD morphology in rab-10 mutants revealed an additional anterior-posterior patterning defect that has not been described previously . While branches are lost from the posterior dendrite , more exuberant branching is observed in the anterior distal dendrite compared with wild-type dendrites . This anterior shift of the dendritic arbor indicates that RAB-10 is not merely responsible for branch outgrowth but also for regulating the distribution of branching activity along the dendrite , a process we hypothesized to be mediated by molecular motors . Indeed , we found dendrite morphology defects in both dynein and kinesin-1 mutants . Based on the MT polarity in each genotype and their corresponding dendritic morphology phenotypes , we propose that RAB-10 potentiates UNC-116/kinesin-1 mediated transport , which moves cargoes from the anterior to posterior dendrite . This model is further supported by the observed phenotype in the unc-116; rab-10 double mutant , which showed a suppression of the posterior branching defects seen in rab-10 . Our work indicates that PVD achieves its stereotyped dendritic branching pattern by balancing anterograde and retrograde transport of branching factors , which requires the correct establishment of microtubule polarity , the appropriate balance of motor activity , and the correct association of transport cargoes with these motors . We propose a model in which RAB-10 is required for this balance of motor activity . First , microtubule polarity in the anterior and posterior primary dendrites dictates that kinesins , such as UNC-116 , traffic membrane organelles from the anterior to posterior dendrite . Dynein does the opposite . In wild-type animals , these two motors have a particular equilibrium of activity , which results in the “normal” distribution of branching factors . Second , the kinesin activity is slightly stronger than the dynein activity in PVD , which sets up a subtle posterior-anterior gradient with fewer branches in the distal anterior dendrite . Third , RAB-10 promotes kinesin-1 mediated trafficking . In the rab-10 mutants , this model predicts that kinesin-1-mediated transport of branching factors is reduced , and dynein activity dominates , resulting in a shift of the distribution of branching factors towards the anterior dendrite . This shift is dependent on microtubule polarity , as demonstrated by the unc-116; rab-10 double mutant , where microtubule polarity is reversed . In this mutant , dynein functions as a retrograde motor , which rescues dendritic branches in the posterior region . We demonstrate a genetic interaction between RAB-10 and kinesin-1 , and a physical interaction has been previously shown in hippocampal axons , where Rab10 mediates anterograde trafficking by associating with kinesin-1 and its adaptor protein JIP1 [21] . In PVD , RAB-10 may similarly regulate the balance of motor activity by directly activating UNC-116 . Another possibility is that RAB-10’s role in balancing motor activity arises from a preferential association of RAB-10-positive plasmalemmal precursor vesicles with kinesin over dynein or that RAB-10 is involved in the appropriate sorting of kinesin and dynein cargos . Additionally , it has previously been shown that reduction of either dynein or kinesin-1 activity in Drosophila sensory neurons results in the loss of distal dendrites , suggesting that across many species , balancing motor activity is necessary to distribute branching components across developing dendrites [47] . Future research will focus on further characterization of the relationship between RAB-10 and the molecular motors to establish the nature of the genetic interactions between RAB-10/UNC-116 and RAB-10/dynein in the growing PVD dendrite . In addition to RAB-10’s role in patterning the PVD dendrite , we have also demonstrated that RAB-10 is required for branch outgrowth and the subcellular localization of transmembrane proteins such as DMA-1 . This suggests a second role for RAB-10 , in which RAB-10 and the exocyst coordinate docking and secretion of membrane and protein cargoes . RAB-10 has previously been shown to work in conjunction with the exocyst complex in both endocytic trafficking and tracheal outgrowth [23–25] . In the PVD dendrite , we found that the loss of two exocyst complex proteins , EXOC-8 and SEC-5 , results in reduced dendritic complexity in the posterior regions of PVD , nearly phenocopying rab-10 . We also found that the constitutively active RAB-10 ( GTP ) was able to rescue the dendritic phenotype of exoc-8 , suggesting that rab-10 and exoc-8 act in a common pathway to support dendritic branch formation . Additionally , we demonstrated that a transmembrane protein responsible for dendrite growth and branching , DMA-1 , was disrupted in rab-10 mutants . These results suggest that RAB-10 and the exocyst complex play a role in the secretion of proteins that control dendritic outgrowth and branching . The requirement of RAB-10 and the exocyst for dendritic outgrowth and membrane protein localization were also described in a recent report [38] , which adds independent support to our discovery that RAB-10 and the exocyst are required for dendritic branch morphogenesis . We therefore propose that RAB-10 and the exocyst act together to control dendritic outgrowth by supporting docking and exocytosis of protein and membrane cargoes at appropriate subcellular dendritic domains . Taken together , our data provide support for new roles for the small GTPase RAB-10 in dendritic outgrowth and patterning . RAB-10 coordinates the distribution of branching factors along the A-P axis of the growing PVD dendrite , and in conjunction with the exocyst complex , RAB-10 allows for the docking and secretion of branching factors such as DMA-1 . Our work provides novel insight into the regulation of branch distribution by RAB-10 and demonstrates that the regulation of dendritic trafficking by Rab GTPases is of critical importance in the establishment of complex dendritic arbors .
Worms were raised on OP50 Escherichia coli-seeded nematode growth medium plates at 20°C [48] . The wild-type reference strain was N2 Bristol . DV2689 sec-5 ( pk2358 ) /mln[dpy-10 ( e128 ) mls14] , RB1928 exoc-8 ( ok2523 ) , and VS1026 rab-10 ( ok1494 ) were obtained from the Caenorhabditis Genetics Center . The wy787 allele was isolated from an F2 semiclonal screen of 3 , 000 haploid genomes in the PVD::myrGFP strain . Worms were mutagenized with 50mM ethyl methanesulfonate . The rab-10 locus was identified with SNP mapping , fosmid rescue , and sequencing . Two PVD-specific promoters , ser2prom3 and pdes2short [30] , were used interchangeably for PVD specific expression; phlh-1 was used for muscle-specific expression . The integrated wyIs592 ( ser2prom3::myrGFP ) and the extrachromosomal array ser2prom3::myrmCherry were used to visualize PVD . wyEx4286 ( ser2prom3::dma-1::GFP ) was used to visualize DMA-1 localization . wyEx4985 ( pdes2::ebp-2::GFP ) was used to visualize plus-end microtubules . Transgenes generated for this work include: wyEx7698 ( ser2prom3::rab-10 ) , wyEx7701 ( pdes2short::rab-10::GFP ) , wyEx7700 ( pdes2short::rab-10 ( GTP ) ::GFP ) , wyEx7772 ( pdes2short::rab-10 ( GDP ) ::GFP ) , wyEx7763 ( ser2prom3::exoc-8 ) , wyEx7699 ( hlh-1::rab-10 ) , rab-10::rab-10 , wyEx7069 ( PVD::mcd8 . 3 ) , ser2prom3::rab-10::mCherry; wyEx7137 ( ser2prom3::rab-7::mCherry ) , wyEx50081 ( Phsp16 . 2::Cas-9;PU6::dhc-1-sgRNA ) , and Phsp16 . 2::Cas-9;PU6::denn-4-sgRNA . Podr-1::RFP and Punc-122::RFP were used as coinjection markers . Images were captured in live animals using a Plan-Apochromat 63x/1 . 4 objective on a Zeiss LSM710 confocal microscope or using a Plan-Apochromat 63x/1 . 4 objective on a Zeiss Axio Observer Z1 microscope equipped with a Yokagawa spinning disk head . For still images , worms were immobilized using 10 mM Levamisole ( Sigma-Aldrich ) in M9 on 2% agarose pads . For FRAP imaging of DMA-1::GFP and time-lapse imaging of EBP-2::GFP , worms were soaked in 0 . 1% tricane/0 . 01% levamisole in M9 for 20 minutes and were then mounted on 5% agarose pads with 0 . 05μg of polysterene microspheres ( Polysceinces ) . Video for EBP-2::GFP comet analysis was acquired at 8 frames per second . Images were straightened with the ImageJ software ( US National Institutes of Health ) using the primary dendrite as a reference . In L4 animals , the PVD dendrite was divided into four segments ( three equal-length anterior segments , and one posterior segment ) and the number of secondary and quaternary branches per segment was counted . Quaternary branches were assigned to a segment based upon the location of the secondary branch forming the base of the menorah from which the quaternary branch originated . Any protrusion from the primary dendrite was classified as a secondary branch , and any correctly-oriented protrusion from the tertiary dendrite was classified as a quaternary branch . A “branch complexity” index , defined in Fig 1 , was computed for each segment . This term weights each branch order and is normalized to an “ideal” segment . In L2 and L3 animals , only the total number of secondary branches was tabulated . CRISPR-Cas9 constructs were designed according to previously established methods [40] . Phsp16 . 2 , a heat-shock promoter expressed in many tissues , was used to drive Cas9 expression . Worms were synchronized via bleaching and L1 plates were heat shocked at 33°C for two hours . The efficiency of the CRISPR-Cas9 editing was determined by observing >30 animals of each genotype . For the denn-4 experiment , which had ~20% efficiency , animals with the CRISPR-Cas9 array and an observed PVD phenotype were selected for imaging and quantification , in comparison to wild-type controls without the array from the same plate . For the dhc-1 experiment , all animals with the CRISPR-Cas9 array were included for imaging and quantification , without pre-screening for neuronal phenotype , because the dhc-1 sgRNA had a much higher efficiency ( ~60% ) . Animals from multiple replicates of each CRISPR experiment were pooled for the quantification . FRAP experiments were performed using the Zeiss Axio Observer Z1 microscope equipped with a Yokagawa spinning disk head . A small region of primary dendrite just anterior to the cell body was bleached for 1000ms . Recovery images were taken every second for five minutes . Branch complexity numbers per segment were compared across all genotypes using the 2-way ANOVA with post-hoc Tukey’s multiple comparisons test . Similarly , branch complexity numbers were compared across all genotypes within each somatic CRISPR experiment using the 2-way ANOVA with post-hoc Tukey’s multiple comparisons test . FRAP data were compared using 2-way ANOVA . Developmental differences in secondary branches between wild-type and wy787 were assessed using multiple t-tests . | Building a complex dendritic arbor requires tremendous cellular growth , and how membrane and protein components are transported to support a rapidly growing , polarized dendrite remains unclear . We have identified the small GTPase RAB-10 as a key regulator of this process , providing insight into both dendritic development and the control of trafficking by small GTPases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | RAB-10 Regulates Dendritic Branching by Balancing Dendritic Transport |
High throughput sequencing has facilitated a precipitous drop in the cost of genomic sequencing , prompting predictions of a revolution in medicine via genetic personalization of diagnostic and therapeutic strategies . There are significant barriers to realizing this goal that are related to the difficult task of interpreting personal genetic variation . A comprehensive , widely accessible application for interpretation of whole genome sequence data is needed . Here , we present a series of methods for identification of genetic variants and genotypes with clinical associations , phasing genetic data and using Mendelian inheritance for quality control , and providing predictive genetic information about risk for rare disease phenotypes and response to pharmacological therapy in single individuals and father-mother-child trios . We demonstrate application of these methods for disease and drug response prognostication in whole genome sequence data from twelve unrelated adults , and for disease gene discovery in one father-mother-child trio with apparently simplex congenital ventricular arrhythmia . In doing so we identify clinically actionable inherited disease risk and drug response genotypes in pre-symptomatic individuals . We also nominate a new candidate gene in congenital arrhythmia , ATP2B4 , and provide experimental evidence of a regulatory role for variants discovered using this framework .
Since the completion of the human genome project , technological advances have dramatically increased throughput and decreased the cost of human DNA sequencing[1] , facilitating comprehensive interrogation of coding regions of the genome , transcripts , and whole genome sequences . High throughput sequencing has illuminated the underlying genetic basis for rare inherited disease syndromes[2–4] , refined our molecular understanding of cancer pathogenesis[5] . provided a fine map of rare genetic variation underlying common disease risk[6–9] , and refined clinical diagnosis and medical therapy[10–13] . These initial advances and the continued drop in the cost of high-throughput sequencing have prompted predictions of a new era of medicine personalized to individual genetics . However , downstream interpretation of sequence variation data remains a formidable barrier to full realization of the promise of genomic medicine , whether it be applied for investigating the genetic basis for well-described disease phenotypes in individuals and families or for prognostication of disease risk and drug response[1] . Several applications and data resources exist for predicting the effects of genetic variation on human phenotypes[14–17] , but there does not yet exist a comprehensive , widely accessible application for interpretation of whole genome sequence data . We previously developed and applied a methodology for interpretation of genetic and environmental risk in a single participant using a combination of traditional clinical assessment , whole genome sequencing , and integration of genetic and environmental risk factors[18] , and extended this framework to familial context[19] . Here we describe an integrated pipeline , Sequence To Medical Phenotypes ( STMP ) for interpreting high-throughput human DNA sequence data . STMP performs targeted genotyping of variants with known clinical associations , rich functional annotation of discovered variants , and prioritizes genetic variants according to potential impact , mode of inheritance , and phenotypic presentation . For individual genome sequences , STMP provides predictive genetic information regarding risk for inherited disease traits and response to pharmacological therapy . We demonstrate the use of this analytical pipeline for disease and drug response prognostication in pre-symptomatic individuals , and for elucidation of the genetic basis congenital ventricular arrhythmia .
This study was reviewed and approved by the Research Compliance Office at Stanford University ( protocol # 4237 , SQL 96726 ) . Informed written consent was obtained from all participants . Methods for whole genome sequence ( WGS ) interpretation in the context of disease gene finding in inherited disease syndromes and predictive genetic variant annotation are outlined in heuristic fashion in Fig 1 . High-throughput re-sequencing currently requires a reference genome for sequence assembly and variant identification . The reference genome that is currently used for alignment of human re-sequencing data and variant identification ( the NCBI reference genome ) [20] is derived from a collection of DNA samples from a small number of anonymous volunteers . However , it represents a very small sampling of human genetic variation . As such , at ~1 . 6 million genomic positions , the NCBI reference sequence differs from the major allele in each of the three Haplotype Map ( HapMap ) populations . These minor alleles span ~4 , 500 variant positions associated with common complex disease and drug response traits[19] , including the Factor V Leiden allele associated with hereditary thrombophilia . Comparison of genome sequence data that is homozygous for these alleles to this reference sequence will naturally not produce a variant call . This issue is partially addressed by the use of major allele reference sequences[19] . A more comprehensive approach is to perform targeted genotype calling of all loci considered to be of phenotypic importance . This approach is reference agnostic up to reference base bias associated with short read alignment and mapping . STMP uses input interval call files representing previously reported Mendelian disease associated loci and loci associated with drug response to provide targeted genotype calls , irrespective of reference base , using the GATK Unified Genotyper ( for SNVs ) and Haplotype Caller ( for indels ) and their capability to report genotypes and coverage for the “reference calls” . STMP also provides metrics for coverage of loci with known importance to human health and disease , thereby providing confidence that , for example , a given disease-associated allele is indeed not present , rather than just under-sequenced or otherwise not confidently ascertained . As compared with methods that store diploid calls for all reference genomic positions , the STMP approach to genotype interrogation facilitates downstream variant annotation while minimizing storage requirements for genotype data . To facilitate updated genotype identification as new loci of relevance to human health and disease are discovered , binary alignment ( BAM ) files from the secondary sequence analysis are retained for future use . Rich functional genomic annotation is a prerequisite to sequence interpretation pipelines that aim to provide testable biological hypotheses about the basis for described disease syndromes and for disease risk and drug response prognostication . We extended the annovar framework [21] to provide rich gene-based , functional genomic , regulatory , allele frequency , and phenotypic annotation . This basic annotation pipeline provides 94 annotations for SNVs and indels in VCF format ( http://www . 1000genomes . org/wiki/Analysis/Variant%20Call%20Format/vcf-variant-call-format-version-41 , S1 Table ) and 39 annotations for SVs in GFF format ( https://www . sanger . ac . uk/resources/software/gff/spec . html , S2 Table ) . STMP can also leverage gene co-expression network topology information to provide quantitative prior expectations about gene-level pathogenicity for contextualizing individual variation data . For example , STMP output may be used to identify genetic variation occurring in genes that are co-expressed with known disease genes , thereby implicating by association variants perturbing certain network topologies . The default STMP module comes pre-loaded with gene co-expression network topology representing gene expression microarray data from 75 normal unused human donor hearts , tissue from 49 human hearts with right- or left-ventricular hypertrophy , and 436 explanted human hearts with dilated cardiomyopathy . The general framework for weighted gene coexpression network analysis is described [22 , 23] . Brielfy , pair-wise Pearson’s correlation between gene expression values was calculated for every gene in the dataset for: a ) samples from normal unused donor hearts , b ) samples from hearts with right- or left-ventricular hypertrophy , c ) samples from hearts with dilated cardiomyopathy . A soft-thresholding parameter β was chosen to satisfy scale-free topology criterion based on r2 maximization for a linear fit with slope –1 to log ( k ) versus log ( n ( k ) ) , and the topological overlap between genes was calculated [24] , generating a network adjacency based on shared network neighbors . We next used average linkage hierarchical clustering and the dynamic tree cut algorithm [25] , to partition the topological overlap network into modules . Disease-specific topologies can be used to assess dynamic gene-gene interactions that are context specific . Genome-wide genetic interrogation in father-mother-child trios with apparently simplex phenotypes can be a powerful tool for genetic association discovery . Classically these investigations are performed using discrete filtering to identify apparent de novo , compound heterozygous , and rare homozygous mutations . De novo mutations are typically discovered via searching for Mendelian inheritance abnormalities ( MIAs ) that are consistent with the segregation of phenotypes within the family . Discrete filtering is encumbered by several challenges , however . First , the true per-generation de novo mutation rate [26 , 27] is two to three orders of magnitude higher than the sequence error rate using current high throughput sequencing technology [27 , 28] . In addition to stochastic error modes , there are systematic error modes that relate to sequences in the human reference genome that are compressed relative to common repetitive sequences , low complexity and GC and homopolymer rich regions , and other regions of the human genome that are problematic to accurately sequence , align , or genotype . Roach , et al , developed an HMM-based classifier to identify these regions in family quartets and also to provide relative inheritance state information for WGS [27 , 29] . STMP utilizes a simplified hidden Markov model ( HMM ) classifier that bins WGS or WES data into one of three categories: “good data” , “compression” , or “MIA-rich” regions . The latter two categories represent variant data that is highly likely to contain systematic artifact and can be excluded from downstream analysis and-or interpretation . STMP writes this information in the “Filter” field of standard VCF output , allowing for soft-filtering or manual inspection of these regions . STMP uses chromosomal distance between variant markers as a prior expectation in the HMM , thereby facilitating the use of this HMM-based approach in WES , which by virtue of sequence capture is sparse outside targeted regions and dense within targeted regions . Second , apparently simplex disease phenotypes can arise from a variety of possible underlying genetic architectures and modes of inheritance , and pre-specifying one mode can lead to a lack of sensitivity . To address this , once HMM-based regional classification has been performed , STMP will output 1 ) apparent de novo events , 2 ) all instances of compound heterozygosity in a gene in which at least one variant in the pair is rare according to user specific criteria , 3 ) rare homozygous mutations , and 4 ) instances of apparent hemizygosity which are candidates for loss of function , 5 ) rare variants in known inherited disease genes fitting an autosomal dominant inheritance model with reduced penetrance . This output can be used for manual inspection in single trio studies or as a prior expectation for gene regions fed forward into collapsing statistics if a cohort of trios is studied . In the latter case STMP leverages inheritance information to reduce the number of gene regions queried and thus the number of statistical comparisons performed between case and control cohorts . Predicting drug response from WGS data requires generation of best-guess haplotypes from short-read sequence data for which haplotype phase is often not determined molecularly . STMP produces best-guess haplotype pairs from confidently genotyped SNVs and indels identified as above . To do this STMP first creates skeleton haplotype pairs using all confidently identified homozygous SNVs . The full set of complementary haplotypes is then generated using heterozygous variant calls . A perfect-match search is performed for each haplotype and its complement among described haplotypes defining the known “star” alleles associated with clinical drug response . If a perfect match is not found , the set of possible haplotype pairs is given but no star allele assignment is attempted . If more than one pair of possible star alleles is found matching possible haplotypes generated from WGS data , all possible star allele combinations are reported . STMP does not provide haplotype resolution beyond that suggested by the confidently called genotypes . That is , if a variant is not confidently called or not covered , as may be the case in exome or other targeted sequencing , haplotypes that are uniquely defined by these “tags” variants are not disambiguated from other possible star allele-defining haplotypes , and a set of possible star alleles corresponding to each reduced haplotype and its conjugate are reported . The haplotype determination is purposefully designed to only give high-confidence predictions , leaving the task of disambiguating star alleles in the setting of uncertain genotype calls or uncommon haplotypes to a human curator . STMP also annotates and reports single variant drug response associations cataloged in the PharmGKB knowledgebase [30] at a level of evidence ( for definitions , see http://www . pharmgkb . org/page/clinAnnLevels ) defined by the user ( Fig 1 ) . Following genotype annotation , STMP prioritizes variants by using metrics of conservation and constraint , predictions of pathogenicity , and allele frequency derived from comparisons with local and external data sources . STMP uses a prioritization scheme that at once provides a parsimonious set of candidates for manual review and a comprehensive assessment of previously reported genetic variation . The heuristic for prioritization of previously reported variants in monogenic disease genes , as well as rare and novel variants in monogenic disease genes with no previously reported phenotypic association , is described in Fig 1 . In default mode STMP first reduces all variants to a set that occurs within 2 , 725 genes cataloged in ClinVar ( 2 , 716 in females due to cataloged disease associations within nine genes on the Y chromosome ) [31] , manually curated to exclude drug response associations and common disease susceptibility loci . Alternatively , gene sets can be provided by the user and utilized for genetic variant filtering based on the phenotypic features of the disease queried . Previously reported variants in disease mutation catalogs include a significant number of common polymorphisms , mapping errors , legacy coordinates , and common disease susceptibility loci that are unlikely to be relevant to monogenic disease risk [32 , 33] . Thus , STMP prioritizes variants previously reported in the Human Gene Mutation Database ( HGMD ) first by expected pathogenicity and next by allele frequency . Previously reported variants cataloged in HGMD are separated into four tiers of potential pathogenicity: Variants meeting criteria for tiers 1–3 are retained for downstream manual review in the case of individual genome interpretation , and for intersection with inheritance state information in the case of disease-targeted analyses . As the expected allele frequency of potentially pathogenic variants is likely to vary greatly with disease prevalence , penetrance , and mode of inheritance , allele frequency filters are not used for tiers one and three , thereby allowing for prioritization of functional alleles with previously described disease associations that would not otherwise pass strict allele frequency filters , for instance the deltaF508 allele in CFTR or the Factor V Leiden allele . Rare and private variation , as a result of recent population expansion and purifying selection , continues to constitute a significant proportion of human genetic variation , even in large population surveys . Some of these rare and private alleles will have monogenic disease risk and carrier status relevance , and therefore STMP also prioritizes select previously unreported , but potentially pathogenic , rare and novel variants in monogenic disease genes , incorporating consensus evidence for evolutionary constraint/conservation and pathogenicity prediction . These computational methods for scoring genetic variants have , in isolation , modest classification accuracy and inter-algorithm concordance [35 , 36] . Approaches to rating the potential pathogenicity of variants based on consensus of commonly used prediction algorithms have been shown to have superior calibration and discriminative accuracy when compared with individual predictions [36] . STMP imposes a prior expectation that pathogenic alleles are more likely to occur in genes in which previously reported variants have produced Mendelian disease phenotypes , but also archives and categorizes all other variants for review and potential reclassification as genetic knowledge expands , or for intersection with inheritance state data . Rare ( defined as above ) and novel variants with no previously described phenotype association in monogenic disease genes are prioritized into four tiers of potential pathogenicity: As a final filter for both previously reported variants in monogenic disease genes and previously unreported , rare variants in monogenic disease genes , STMP uses catalogs of local allele frequency and genotype information to exclude variants that are observed more frequently than would be expected in the local sequencing environment . This filter serves to identify and exclude variants whose previously unappreciated high allele frequency is a result of false negatives in population genetic surveys or systematic false positives specific to local sequence variant discovery pipelines . Similar filters have proven effective in excluding such systematic artifacts in other contexts [37] . STMP accepts as required input a vcf file prepared by Illumina Isaac , GATK , the Complete Genomics variant discovery pipeline , or Real Time Genomics . Optional inputs include 1 ) binary alignment map file , required for targeted genotype calling and annotation of known inherited disease risk alleles and pharmacogenomic annotation , 2 ) genome feature format file describing structural variant events , 3 ) local site frequency spectrum for filtering of inherited disease risk candidates . In trio mode STMP requires jointly called genotypes and sample identifiers for father , mother , and child . Initial annotation of genetic variants for gene , functional genomic , and clinical phenotypes is performed using python/perl and parallelized using the “multiprocessing” module in python . Processing time for the annotation pipeline scales roughly as 1/n , where n is the number of processors allocated to the task . STMP is implemented in cython/python/shell and also parallelized using the “multiprocessing” modules in python . To demonstrate the utility of the STMP tool , we applied STMP in trio mode to a trio with congenital neonatal arrhythmia , and also to twelve unrelated adult participants ( median age 53 , 6 female , 7 of East Asian ancestry ) recruited from primary care clinics at Stanford University Medical center . The study was approved by the Stanford University Institutional Review Board and all patients gave informed written consent , or , in the case of minors , assent ( Stanford Institutional Review Board GAP approval number 4237 ) . To further explore a novel gene locus implicated in the congenital neonatal arrhythmia trio , we performed in vitro characterization of a putative promoter variant found in trans with a novel protein truncating mutation in ATP2B4 using previously described approaches[38] . The noncoding variant in the 5’ un-translated region of ATP2B4 , rs4600103 , was first investigated in silico using 1000 Genomes haplotype data and ENCODE regulatory datasets for chromatin accessibility and histone modifications . Predicted transcription factor binding sites ( TFBS ) altered by rs4600103 and the linked variant ( r2~0 . 87 ) , rs4951276 ( also present in the patient ) , were determined using TRANSFAC , JASPAR , and MatInspector positional weight matrix ( PWM ) databases . Sequences surrounding each allele were scanned for vertebrate TFBS with a 0 . 9 minimum similarity score cutoff . Identified putative enhancer elements for each SNP were used to generate concatenated oligonucleotides for rs4600103-G/A and rs4951276-T/A , which were annealed at 95 C for 5 minutes and allowed to cool to room temperature . Resulting double-stranded DNA fragments were subcloned into the multiple cloning site ( MCS ) of the pLuc-MCS vector ( Agilent ) , and constructs were confirmed by Sanger sequencing . Luciferase reporter constructs containing respective major and minor alleles of rs4600103 and rs4951276 were transfected , along with renilla luciferase , into HEK 293 and H9c2 cells using Lipofectamine 2000 ( Life Technologies ) , according to the manufacturer’s instructions . Growth media was changed after 5 hours and dual-luciferase activities measured after 24 hours using a SpectraMax luminometer ( Molecular Devices ) . Firefly luciferase activities were normalized to renilla luciferase and expressed as the fold change of the empty vector control .
The trio format is a common familial arrangement in sequencing studies undertaken to uncover the genetic basis for a known disease or to assist in disease diagnosis [39 , 40] . In many such trios , the offspring is the only clearly affected subject ( “simplex” trios ) , proposing several possible modes of phenotypic expression , including recessive or codominant inheritance , autosomal dominant inheritance with reduced penetrance , and a de novo mutation in the proband . STMP functionally annotates and prioritizes such alleles , including all possible instances of compound heterozygosity and important noncoding variants . To demonstrate the utility of STMP for discovery of disease-associated genetic variants in individuals with manifest disease in this format , we used STMP in “trio” mode to investigate whole genome sequence data from a father-mother-child trio with neonatal ventricular arrhythmia . In this trio the offspring was affected by neonatal polymorphic ventricular arrhythmia preceded by ST segment elevation ( Fig 2A . ) . Clinical genetic testing of inherited arrhythmia genes , including deletion-duplication testing , did not uncover disease-causing mutations . STMP identified 25 candidate variants . Among the candidate compound heterozygous variants , a novel nonsense mutation and a common 5’ UTR variant were found in trans in ATP2B4; the latter variant , rs4600103 , was found in an accessible and active chromatin region as determined by ENCODE derived DNAse hypersensitive regions and enrichment for promoter histone modifications ( H3K4me3 ) in human cardiac fibroblasts ( HCF ) and cardiac myocytes ( HCM ) and active histone modifications ( H3K27ac ) in human lung fibroblasts ( Fig 2B ) . ATP2B4 encodes a plasma membrane calcium ATPase that mediates neuronal nitric oxide signaling in cardiac myocytes and directly interacts with a gene , SNTA1 , that has been implicated in hereditary ventricular arrhythmia and sudden , presumed arrhythmic infant death [41–43] . Using TRANSFAC and JASPAR transcription factor binding databases , we identified altered motifs for ELK1 and NFκB transcription factor binding sites ( TFBS ) proximal to the SNP . Using 1000 Genomes data we identified another common variant , rs4951276 ( MAF 0 . 35 ASN , 0 . 09 EUR ) in high linkage disequilibrium ( r2 = 0 . 87 ) with rs4600103 , which may explain some of the regulatory effects . This intronic variant resides in a putative enhancer element , containing the active chromatin mark , H3K27ac , and is predicted to disrupt TFBS motifs for FOXP1 . To interrogate the potential impact of these variants in regulating ATP2B4 expression , we cloned the predicted regulatory elements surrounding each allele into a luciferase reporter construct driven by a minimal promoter , and measured relative transcriptional activity in both HEK 293 and the neonatal rat cardiomyocyte cell line , H9c2 . Interestingly , the minor A allele at rs4600103 was shown to have reduced transcriptional activity compared to the major G allele , whereas both alleles at rs4951276 had similar reporter activities ( Fig 2C ) . These results indicate that rs4600103 may be a functional variant identified at ATP2B4 through altering a putative cis-regulatory element . It remains unclear whether this variant , in combination with the truncating variant in trans , is disease associated . It may be that an as yet unidentified factor such as a trans-acting regulatory element , structural variant on the other allele , or environmental- or gender-specific modifier of the phenotype is at play . This uncertainty highlights one of the challenges inherent to identifying a single likely pathogenic allele in a recessive disease gene . When run on WGS data from a single proband , STMP provides , for the first time , pharmacogenomic haplotype assignment and annotation of clinically associated pharmacogenomics alleles , including those that are defined on a single variant or haplotype bases . To demonstrate the utility of STMP in this context , we assessed the concordance between star allele assignments generated by STMP for five genes with associated Clinical Pharmacogenomics Implementation Consortium ( CPIC ) guidelines for drug dosing and administration ( CYP2C9 , CYP2C19 , CYP2D6 , VKORC1 , and SLCO1B1 ) . Haplotype call concordance between STMP and blind manual haplotype determination demonstrated that in all twelve individuals , the star allele pair assigned by human rates was found in the set of possible star alleles reported by STMP for all five genes . As described [44] , STMP provided 1–3 recommendations per subject for change in drug dose or administration , and 3–10 additional high-confidence genetic drug response predictions from WGS data . When applied to WGS data from single probands , STMP provides rich functional annotation and prioritization of potential Mendelian disease risk alleles , including novel variants , structural variants , and important regulatory variants . STMP allows for genome-wide search for such genetic variants , or can be restricted to specific gene sets if a targeted diagnostic question is pursued . To demonstrate the utility of STMP to discover such variants genome-wide , we applied STMP to Illumina WGS sequence data ( median haploid read depth 51x , 101bp x 2 paired end reads , generated on the HiSeq 2000 ) from twelve unrelated adult participants ( median age 53 , 6 female , 7 of East Asian ancestry ) recruited from primary care clinics at Stanford University Medical center . Methods for identification of single nucleotide variants , indels , and structural variants are described in Dewey , et al ( 2014 ) [44] . On a six-core Intel Xeon X5670 processor running 64-bit linux with 128 GB of RAM and utilizing five concurrent threads , stanovar performed comprehensive annotation of standard . vcf and . gff format variant files in a median of 96 ( range 90–102 ) minutes per genome . STMP performed prioritization of Mendelian disease risk candidates and identification of genotypes and haplotypes affecting drug response in < 5 minutes per participant . The median total processing time , including targeted genotype calling of SNVs and indels with clinical associations and filtering based on local site frequency spectra , was 122 minutes ( range 116–127 minutes ) . We used allele frequency filters of <1% in general population surveys and <25% in our local cohort; higher allele-frequency cutoffs for local sequence data may be appropriate in populations enriched for Mendelian phenotypes and associated variant alleles . Higher allele-frequency cutoffs for local sequence data may be appropriate in populations enriched for Mendelian phenotypes and associated variant alleles . Manual curation uncovered several well-established disease causing mutations in this cohort without apparent Mendelian disease , including a 19-bp insertion-deletion variant in BRCA1 that has been previously implicated in hereditary breast and ovarian cancer , prompting prophylactic surgery [44] . Variants discovered in each participant , prior to and after allele frequency filtering , are presented in Table 1 . Further filtering of variants occurring at high allele frequencies in the cohort was particularly effective at reducing the number of previously reported Mendelian disease risk candidates and the number of apparently rare ( according to external allele frequency information ) loss-of-function variants in Mendelian disease genes . This suggests that even a small number of local “control” genomes can substantially reduce the number of potential false positives resulting from systematic sequencing artifact related to the local peculiarities of sequencing and analysis or previously unappreciated common variation .
Here we describe a series of methods for annotation of high-throughput sequence data for individual genetic risk prediction and prediction of drug response . This collection of tools is flexible , customizable , and allows for dynamic interaction between variant annotation and association efforts . It is applicable to variant data from whole genome , exome , and targeted re-sequencing . We further demonstrate that application of these methods to whole genome sequence data in apparently unrelated individuals yields a parsimonious set of variants for manual review ( ~100 ) that may have implications for Mendelian disease risk and drug response , and in one case uncover a clearly clinically actionable disease-causing mutation in a pre-symptomatic individual . Application to a father-mother-child trios uncovered a novel candidate disease gene in neonatal ventricular arrhythmia . While several methods exist for predicting pathogenicity of sequence variants , and other methods exist for annotating variants with respect to described disease associations , there is not yet a unified framework that integrates the bulk of human disease-genotype associations and computational predictions . The set of methods we described here is developed to do just that . Furthermore , in contrast to existing tools that perform limited annotation of SNVs and indels , our integrated pipeline provides a framework for interpreting structural variants and variants disrupting important noncoding regions of genes associated with disease phenotypes . This set of methods leads naturally into manual curation of discovered variants for research efforts utilizing disease phenotype and drug response information . Sequence to medical phenotypes ( STMP ) is an open source , parallelized pipeline for clinical interpretation of WGS and WES data generated in a research setting . It is highly amenable to parallel processing architecture , produces parsimonious variant sets for manual review , and interrogates both Mendelian disease risk and genetic drug response . We hope that the methods presented here will help catalyze future clinical research using WGS . STMP is open source and will be available at http://ashleylab . stanford . edu/tools/stmp . html . | Technological advances have dramatically reduced the cost of sequencing the human genome . Tools for analyzing such data across families including annotation of clinically important variants and aggregation of variants for personalizing drug prescriptions have been developed but few are publically available . Here we describe such tools then demonstrate their application in several distinct data sets . In particular , we use the tools to define the genetic basis of a new congenital arrhythmia syndrome . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Sequence to Medical Phenotypes: A Framework for Interpretation of Human Whole Genome DNA Sequence Data |
The best measure to limit spread of contagious diseases caused by influenza A viruses ( IAVs ) is annual vaccination . The growing global demand for low-cost vaccines requires the establishment of high-yield production processes . One possible option to address this challenge is the engineering of novel vaccine producer cell lines by manipulating gene expression of host cell factors relevant for virus replication . To support detailed characterization of engineered cell lines , we fitted an ordinary differential equation ( ODE ) -based model of intracellular IAV replication previously established by our group to experimental data obtained from infection studies in human A549 cells . Model predictions indicate that steps of viral RNA synthesis , their regulation and particle assembly and virus budding are promising targets for cell line engineering . The importance of these steps was confirmed in four of five single gene overexpression cell lines ( SGOs ) that showed small , but reproducible changes in early dynamics of RNA synthesis and virus release . Model-based analysis suggests , however , that overexpression of the selected host cell factors negatively influences specific RNA synthesis rates . Still , virus yield was rescued by an increase in the virus release rate . Based on parameter estimations obtained for SGOs , we predicted that there is a potential benefit associated with overexpressing multiple host cell genes in one cell line , which was validated experimentally . Overall , this model-based study on IAV replication in engineered cell lines provides a step forward in the dynamic and quantitative characterization of IAV-host cell interactions . Furthermore , it suggests targets for gene editing and indicates that overexpression of multiple host cell factors may be beneficial for the design of novel producer cell lines .
Influenza A viruses ( IAVs ) are highly contagious respiratory pathogens that constitute a permanent threat to public health , causing three to five million cases of severe illness and up 650 , 000 deaths per year [1] . As obligate intracellular parasites , influenza viruses rely on host cellular functions at every step of their life cycle . Thus , to deepen the understanding of virus-host cell interactions is a key step to improve vaccine production and thereby efficiently counteract disease . During the past decade multiple RNAi screens , yeast-two-hybrid approaches and omics studies , allowed for systematic identification of cellular factors that are relevant for the IAV life cycle ( recently reviewed by [2] ) . These factors are commonly grouped into pro- and antiviral factors , which can be used to design new therapeutic and preventive disease measures . So far , the focus of these investigations was mainly on novel antiviral treatment that targets host dependency factors instead of viral factors , which might help to avoid the emergence of viral escape mutants [3–6] . Regarding the design of cell lines for optimized virus production , however , host restriction factors , e . g . factors that belong to cellular antiviral defense mechanisms and which can be downregulated to increase the virus yield in vaccine manufacturing , are of key importance . In the case of poliovirus , for instance , the knockdown of host cell factors that inhibit virus replication in adherent Vero cells was reported to result in a ten-fold increase in virus titers [7] . This promising result , however , could not be reproduced in a recent follow-up study [8] . Another option , pursued in our study , is the overexpression of host dependency factors to facilitate virus replication and increase yields in cell culture-based IAV production . To this end , we chose the lung carcinoma cell line A549 as a model cell line that was previously used in two genome-wide RNAi screens for identification of antiviral targets [9 , 10] ( for further review of relevant RNAi screens the reader is referred to [11] ) . In these studies , changes in virus replication were measured in cells with temporal modulation of gene expression and evaluated at single time points post infection ( p . i . ) . To complement this approach , we investigate the dynamics of virus replication in cell lines stably overexpressing host cell genes over an extended period . Since virus-host cell interactions display highly complex dynamics , mathematical modeling approaches are crucial to support the interpretation of time courses of viral components measured in experiments , e . g . intracellular viral RNA copy numbers . In addition , such models help to explain specific steps and outcomes of virus-host cell interaction , to study effects of changes in expression of viral or cellular components , or to make predictions about phenotypic changes after cell line engineering , i . e . , inhibition of virus growth or increase in yield . We employed a model of the IAV life cycle that describes virus replication within a single infected adherent MDCK cell [12] . First , we re-calibrated this model to experimental data from infected A549 cells obtained in this study . Second , we predicted which steps of the virus life cycle are most sensitive with respect to cell-specific virus yield and therefore represent promising targets for cell line engineering . To validate model predictions , we integrated various experimental data sets from infection studies performed in A549 cell lines that we modified genetically to overexpress host cell factors previously identified by RNAi screening [9 , 13–15] and studies performed by other research groups [16–19] . Finally , the resulting parameter sets for IAV replication in single gene overexpression cell lines ( referred to as SGOs ) , were used to predict the outcome of IAV infection in multiple gene overexpression cell lines ( referred to as MGOs ) . While only one of five of the selected SGOs showed a higher virus yield compared to the parental A549 cell line , MGO simulations indicated that there is a potential for a significant increase in virus yield . However , this finding was confirmed only partially in experiments . Overall , SGOs and MGOs that were established during this study showed an improvement in early release dynamics rather than the expected increase in total virus yield compared to their parental cell line . Using a single cell model of IAV replication , we elucidate this in greater detail and link the overexpression of host cell factors to changes in key parameters of virus growth , which has not been reported before .
The model of IAV replication used in this study is identical to a previously published description of the intracellular life cycle of IAV [12] . In general , we assume that basic mechanisms of IAV replication are similar in different host cell lines , but that values for key parameters of virus growth have to be adapted for each host cell system . While the previous model [12] was calibrated against various experimental data , mostly acquired from infected MDCK cells [20 , 21] , the re-calibration of the model used in this study was based on three sets of in-house experimental data from infected A549 cells ( S1 Fig ) . The available measurements allowed to estimate the kinetic parameters for nuclear import of vRNPs kImp , the synthesis of viral mRNA , cRNA and vRNA ( kMSyn , kCSyn , kVSyn ) as well as binding of matrix protein 1 ( M1 ) kM1Bind and the release of viral progeny kRel . Statistical testing ( Table 1 ) revealed that kCSyn and kM1Bind were not significantly different in A549 compared to MDCK cells [12] . However , kVSyn was significantly increased and kMSyn significantly reduced in A549 cells , respectively . Two simplifying assumptions were made to simulate the influence of host cell factors on IAV replication . First , we considered that each step in the virus life cycle was dependent on one host cell factor and secondly , that a change in the expression level of this host cell factor would directly translate into a change of the corresponding kinetic parameter value in our mathematical model for IAV replication . For instance , if a host cell factor responsible for vRNA synthesis is overexpressed , vRNA replication is enhanced , resulting in a higher vRNA synthesis rate . Likewise , the downregulation of the same factor would result in a reduced vRNA synthesis rate . Based on these assumptions , we performed in silico engineering of A549 cells by perturbing each parameter of our model individually with the objective to maximize virus yield at 24 h p . i . ( optimized parameter values are summarized in S1 Table ) . By comparing the simulated virus release of parental A549 cells to results obtained for in silico optimized cell lines ( Fig 1 ) , we observed three possible outcomes upon parameter perturbation: ( i ) virus release dynamics were not affected significantly , ( ii ) only onset of virus release was improved , starting at least 1 h earlier compared to the parental A549 cell line and ( iii ) virus release dynamics were affected significantly leading to an increase in final yield by at least two-fold . The latter was caused by perturbations of parameters that define the most promising targets for cell line engineering , namely steps of viral RNA synthesis , its regulation and virus release ( Fig 1 , green shaded subfigures ) . Interestingly , the model predicted that the upregulation of viral mRNA synthesis is beneficial for virus replication whereas synthesis of viral cRNA and vRNA should be downregulated . To investigate this in greater detail we , next , compared the dynamics of the simulated intracellular viral RNAs and protein levels in both upregulation and downregulation scenarios to levels in parental A549 cells ( Fig 2 ) . We observed that changes of intracellular replication dynamics were most evident upon manipulation of viral mRNA synthesis ( Fig 2 , middle panel ) . Most importantly , the sole increase of the mRNA synthesis rate lead to a higher increase in vRNA levels than the upregulation of the vRNA synthesis rate itself ( Fig 2 , upper and middle panel second column ) . This strongly indicates that viral RNA replication in A549 cells is already saturated and only if more viral mRNA , and consequently , more viral proteins were available , more vRNA could be produced and virus release could be enhanced significantly . In addition , the modulation of regulatory steps , which is accounted for in our model by binding of M1 ( negative regulator ) , had only an impact on final RNA and protein levels rather than on the dynamics per se ( Fig 2 , bottom panel ) . To validate our model predictions , we used lentiviral gene transfer to generate A549 cell populations that overexpress specific host cell genes relevant for IAV replication . The host cell factors CEACAM6 , FANCG , NXF1 , PLD2 and XAB2 were selected from a set of candidate genes determined previously by RNAi screening [9 , 13–15] and virus-host cell interaction studies [16 , 17] . An overview of genes and their function in the IAV life cycle is given in S2 Table . The resulting cell populations were subjected to fluorescence activated cell sorting ( FACS ) to enrich cells that express the transduced gene based on eGFP , which is the co-expressed reporter gene . SGOs that showed stable gene overexpression were infected with A/Puerto Rico/8/34 ( A/PR/8/34 , H1N1 ) at a multiplicity of infection ( MOI ) of 10−4 , which is usually applied for vaccine production processes . We compared virus titers of each SGO to that of the parental A549 cell line at selected time points p . i . ( Table 2 ) . Assuming that changes in virus release are associated with changes in intracellular mechanisms , we selected SGOs for further characterization of intracellular virus replication based on their HA titer . To facilitate selection , we ranked the HA measurements for each time point and each cell line according to their relative increase compared to the parental A549 cell line . As can be seen by the measurement data and the corresponding ranking values in Table 2 , HA titers of all SGOs were increased at early time points p . i . , whereas none of the SGOs showed an increase greater than 20% of the final HA titer at the usual time of harvest 72 h p . i . Thus , by modulating the expression level of these host cell factors , it was possible to influence the IAV release dynamics , however , the total virus yield was similar comparing SGOs to their parental cell line . Next , we performed a detailed characterization of intracellular steps of viral growth in IAV-infected SGOs as well as in the parental A549 cells and an eGFP transduction control ( Figs 3–5 ) . Although only NXF1 SGOs showed a promising increase in virus yield , it seemed that overexpression of host cell factors can influence IAV replication on the intracellular level . Thus , we also explored the possibility whether additive or even synergistic effects on IAV yield could be achieved by overexpressing multiple host cell factors simultaneously . At first , we investigated this option by a computational approach and simulated the virus release of single cells overexpressing different combinations of multiple host cell factors . Since integration of genes into the host chromosome is random , the gene constructs will be inserted at different chromosomal locations with different transcriptional activities and , since transduction follows a Poisson distribution , not every cell will obtain the same number of the gene constructs . Together , these factors influence the strength of overexpression . In addition , the integration process can also have an impact on the gene expression through off-target effects . To account for all these scenarios , which involve some sort of randomness , we used randomized sets of parameters assembled based on the median values of the model parameters kImp , kVSyn , kCSyn , kMSyn , kM1Bind and kRel , previously estimated from experimental data of infected SGOs and the parental A549 cell line . The parameter set of the latter was also included to account for off-target effects . For instance , the parameter set of an MGO may be composed of kImp of XAB2 SGOs , kVSyn of PLD2 SGOs , kCSyn of NXF1 SGOs , kMSyn of FANCG SGOs , kM1Bind of CEACAM6 SGOs , and kRel of the parental A549 cell line . We assume that all transduced genes can be expressed theoretically with the same probability , i . e . , that there is an equal chance that kinetic parameters of the SGOs will be selected during randomization . Note , that even if all five candidate genes were transduced , not every MGO single cell will be a phenotypic mixture of all SGOs , but its parameter set could be kImp and kVSyn of the parental A549 cell line , kCSyn and kMSyn of CEACAM6 SGOs and kM1Bind and kRel of the NXF1 SGOs . To generate in silico MGOs , we chose to randomize parameter sets of those SGOs that showed a beneficial change in parameters according to initial model predictions of this study ( Fig 1 ) . Thus , we combined parameter sets of the top three candidates with the highest virus release rate kRel ( CEACAM6 ( C ) , FANCG ( F ) and NXF1 ( N ) , CFN in Fig 7 ) , the top three with the lowest cRNA synthesis rate kCSyn ( FANCG ( F ) , PLD2 ( P ) and XAB2 ( X ) , FPX in Fig 7 ) , and the top three with the lowest M1 binding rate kM1Bind ( NXF1 ( N ) , PLD2 ( P ) , XAB2 ( X ) , NPX in Fig 7 ) . Finally , we also randomized parameter sets of all SGOs ( CFNPX in Fig 7 ) . In a Monte Carlo approach , we generated multiple randomized parameter sets according to the selected combinations of SGOs and simulated virus infection at MOI 1 for 48 h ( S3 Fig ) . Finally , we evaluated every single cell simulation for the time point at which the first simulated virus particle was released t ( VRel≥1 ) and the fold change in the maximum number of released viral progeny ( Fig 7 ) . Interestingly , these model predictions revealed that a single cell overexpressing multiple genes can theoretically yield up to five-fold more virus progeny than its parental cell line if the underlying parameter set was kImp and kMSyn of the parental A549 cell line , kVSyn of XAB2 SGOs , kCSyn of PLD2 SGOs , and kM1Bind and kRel of the NXF1 SGOs . In particular , the earlier virus release started , the higher was the fold increase in the number of viral progeny . While the time point of first virus release followed a normal distribution , the fold change of virus release showed a log-normal distribution with highly productive cells as rare events . Overall , the combinations CFN , NPX and CFNPX showed similar distributions of the simulation read outs , whereas the combination of FPX resulted in a narrower distribution of virus yield with a slightly lower maximum fold increase of four-fold . Finally , this analysis revealed that highly productive cells are rare events in a heterogenous MGO population and their contribution to the population average is negligible , which leads to an increase of less than two-fold in the final virus yield ( Fig 7 , dashed line in vertical histograms ) . The computational analysis of MGOs indicated that overexpressing multiple host cell factors could result in an earlier onset of virus release and , to some extent , also in an improvement of virus yield . To validate these model predictions , we generated populations of A549 cells in which individual cells express random combinations of selected host cell factors at various levels ( S4 Table ) . In particular , we generated three independent cell populations ( MGO 1-3 ) which provide random combinations of all five host cell factors CFNPX , which also covers the phenotypes of combinations CFN and NPX according to simulations . Further , we generated MGO 4 in which the three factors FPX were randomly combined and which should show a slightly different phenotype compared to CFNPX . All MGOs were infected with IAV at MOI 10-4 . We chose this MOI according to the SGO screening experiment ( Table 2 ) since under these experimental conditions differences between cell lines were more pronounced than for infections at MOI 1 ( Fig 5 ) . Ranking of HA titers revealed that virus release of MGOs was increased at early time points , while final virus yield was not increased significantly in these cell populations compared to the parental A549 cell line ( Table 3 ) . Of note , the impact of overexpressing single host cell genes on virus yield could be enhanced by overexpressing multiple of these host cell genes simultaneously , which partially confirms our model predictions on MGOs . In addition , MGO 4 was the only cell line showing less than 40% increase in virus yield at 42 h p . i . compared to the parental A549 cell line . This supports the model prediction that the combination FPX results in a slightly less productive phenotype than other gene combinations .
IAVs depend on host cellular functions to complete their replication cycle . Our aim was to take advantage of this dependency and manipulate the expression of host cell factors that are relevant for IAV replication to improve virus production for vaccine manufacturing . Due to the complexity of virus-host cell interactions mathematical models are required to complement the interpretation of infection experiments . In the present study , we used a re-calibrated model of IAV replication to predict and quantify changes in virus replication in genetically engineered A549 cells . To account for the influence of host cell factors on steps of the virus life cycle , we made the simplifying assumption that changes in host cell gene expression have a direct impact on kinetic parameters of our model . Although we did not explicitly model physical interactions between host cell factors or cellular pathways with viral components , we were able to identify targets for cell line engineering by evaluating changes in the cell-specific virus release upon parameter perturbations . According to this in silico analysis , both a significant increase in virus yield as well as an earlier onset of virus release could be expected if either viral transcription or translation were significantly enhanced . In contrast , the model predicted that various steps of virus replication need to be downregulated to achieve a higher cell-specific virus yield . For instance , the binding of M1 to nuclear vRNPs , which mediates the nuclear export of vRNPs , should be delayed . The lower the binding rate of M1 kM1Bind , the longer vRNPs serve as template for viral genome replication and transcription inside the nucleus . Accordingly , not only more viral genome copies but also mRNAs will be synthesized and , thus , higher viral protein levels will be achieved ( Fig 2 , lower panel ) , which together will benefit virus yield . Furthermore , the model predicts that a decrease in the vRNA synthesis rate , in the cRNA synthesis rate , and a delayed binding of NP to naked viral RNA , needed to form replication-competent vRNPs and cRNPs , will cause an increase in virus yield ( Fig 1 ) . These three predictions seem counterintuitive since they cause a slowdown of viral replication . On the other hand , however , this strongly suggests that there is an imbalance between viral RNA replication and viral protein synthesis . While the synthesis of viral genomes is saturated , i . e . , the RNA synthesis rates are too high , the supply of viral proteins either needed to form RNPs ( NP and polymerases ) or needed for virus budding ( HA and NA ) represents a limiting step in A549 cells . Interestingly , Ueda and colleagues [23] made similar observations when comparing IAV growth in MDCK and A549 cells . While steps of viral replication were similar in both cell lines , A549 cells released fewer virions because both the maturation of glycoproteins and their transport to the plasma membrane were slower compared to MDCK cells . In line with that , parameter perturbation studies with the single cell model for MDCK cells [12] did not point to bottlenecks in viral transcription and translation ( S4 Fig ) . Indeed , the MDCK-based model is more sensitive to a change in the vRNA synthesis rate compared to a change in the protein synthesis rate , while the A549-based model is highly sensitive to changes in the protein synthesis rate ( S5 Fig ) . We generated cell lines overexpressing host cell genes beneficial for virus replication previously determined by RNAi screening [9 , 13–15] and studies on virus-host cell interactions performed by other research groups [16–19] . Overall , the maximum virus yield was similar in all A549 cell populations . However , the engineered cell populations released more virus particles at earlier time points compared to the parental cell line during infection studies performed at low MOI . To assure that target genes were stably overexpressed , we confirmed the expression of the functionally linked reporter gene coding for eGFP by flow cytometric measurements during cell culture maintenance ( S6 Fig ) . Furthermore , we determined relative expression levels of the transgenes in SGOs by RT-qPCR ( S3 Table ) . Although the overall number of virus progeny produced by engineered cells was not significantly higher compared to the parental cell line , we could not exclude that intracellular mechanisms of virus replication had changed due to the modulation of host cell gene expression . To elucidate this in greater detail we investigated virus replication dynamics on the intracellular level both experimentally and computationally . With the help of the single cell model , we quantified the changes in key kinetic parameters by fitting to the available experimental data . In contrast to our initial model predictions , both nuclear import rate and viral mRNA synthesis rate were reduced in some SGOs compared to their parental A549 cell line . For instance , the viral mRNA synthesis rate in infected cells overexpressing the nuclear export factor NXF1 was only 60% of the one in parental A549 cells , which alone would lead to a reduction in virus yield by 50% . Still , the NXF1 SGO was the only cell line with a higher cell-specific virus yield when infected at MOI 1 ( Fig 5D ) . The model can only capture these experimental data by an increase in the virus release rate . Hence , the improved virus release rescues virus yields such that despite the adverse changes in viral RNA synthesis , the SGOs release equal or slightly higher amounts compared to the parental A549 cell line . It was reported that inhibition of NXF1 in A549 cells impairs nuclear export of viral mRNAs encoding for NP as well as the surface proteins hemagglutinin ( HA ) and neuraminidase ( NA ) [18] . Upon NXF1 overexpression viral mRNA export might be improved , which may lead to an earlier onset of translation , such that viral surface proteins are available earlier compared to the parental A549 cell line , which is less efficient in protein maturation and trafficking [23] . In the single cell model these steps are not explicitly modeled but lumped into a joint release mechanism that depends on the availability of viral proteins and genome copies in the cytoplasm ( S1 File , Equation 27 ) . In addition , the importance of the virus release mechanism was also shown by initial model predictions ( Fig 1 ) that identified virus assembly and budding as kinetic bottleneck of virus production . The overall tendency that an increase in the virus release rate can compensate adverse changes in RNA synthesis steps can also be observed for infected CEACAM6 SGO cells . In contrast to NXF1 , CEACAM6 is not directly involved in steps of RNA synthesis but seems to interact with newly synthesized viral NA proteins during infection , which activates the Src/Akt survival pathway in A549 cells as shown by Gaur and colleagues [16] . In the same study , CEACAM6-silenced A549 cells showed reduced levels of viral genome copies and proteins . However , in our study , the overexpression of CEACAM6 was not beneficial for IAV replication . Accordingly , temporal upregulation of CEACAM6 instead of high abundance seems to be crucial for cellular survival signaling during infection . Furthermore , members of the CEACAM family are already upregulated upon infection by different influenza virus strains , as recently also shown for CEACAM1 and CEACAM5 [24] . In particular , CEACAM1 induction triggers the innate antiviral host cell response by suppression of the translational machinery and limits viral spread [25] . Taken together , the ambivalent role of the CEACAM family and , in particular , the functional role of CEACAM6 in cellular survival pathways , may support the finding that the overexpression of CEACAM6 can be disadvantageous for IAV replication . Still , it is remarkable that CEACAM6 SGO cells release equal amounts of progeny virions compared to parental A549 cells , indicating that despite a certain inhibition of replication , the virus maintains a basal level of reproduction . Except for SGOs NXF1 and CEACAM6 , for which the nuclear import rate was slightly reduced ( p ≤ 0 . 1 , calculated by one-sided Gauss test ) , the nuclear import rate of vRNPs was similar in the other SGOs compared to parental A549 cells . For the PLD2 SGO , this was unexpected , since it is known that inhibition of PLD2 results in delayed virus entry and reduced viral titers [19] . Still , overexpressing PLD2 did neither improve virus entry nor virus release in our study . The only change in kinetic parameters , that was in agreement with initial model predictions ( Fig 1 ) and should benefit virus yield , was the reduction of the cRNA synthesis rate to 50% compared to parental A549 cells . However , this alone would result in an increase of virus yield by only about 1 . 3-fold in simulations , a small improvement that is eliminated by a simultaneous decrease in the mRNA synthesis rate in PLD2 SGOs as determined from the experimental data . The candidate FANCG interacts with the three viral polymerase subunits ( PB2 , PB1 and PA ) and has a direct influence on polymerase activity according to a minigenome replicon assay using a vRNA-like reporter gene [17] . In this particular assay , it was demonstrated that a FANCG knockdown resulted in a decrease of polymerase activity by 50% while overexpression of FANCG showed a three-fold increase in polymerase activity . According to our initial model predictions , FANCG would have been the most promising candidate to improve virus yield , in particular , if the mRNA synthesis rate was increased ( Fig 1 ) . Surprisingly , all viral RNA species showed reduced levels in infected FANCG SGO cells . Although we have only performed two independent experiments to measure intracellular viral RNA levels in infected FANCG SGO cells , RNA copy numbers were lower compared to those in infected A549 cells in the same experiments as well as compared to the averaged RNA levels in A549 cells from all four independent experiments . Taken together , it seems that an overall increase of the viral polymerase activity results in imbalanced virus replication . Therefore , additional simulations were performed to test the effect of increasing all three or different combinations of the RNA synthesis rates simultaneously . However , by only increasing the vRNA synthesis rate , a reduction in virus yield is predicted ( S7 Fig ) , while any other scenario leads to an increase in final yield in simulations ( for instance see S8 and S9 Figs ) . Hence , our experimental observations together with the model-based analysis of this candidate are not in agreement with the study of Tafforeau and colleagues [17] . On the one hand , this may indicate that observations in an ( artificial ) minigenome replicon assay can only give hints towards changes in mechanisms and that the observation in the context of an infection , i . e . , including additional regulatory steps of replication and availability of cellular and viral precursor molecules , can be contradictory . On the other hand , FANCG also has a beneficial function for the host cell , since it is involved in DNA repair mechanisms . We could , therefore , speculate that damage of cellular DNA induced by IAV infection [26] is reduced by overexpressing FANCG . However , we cannot exclude that FANCG plays a pro-viral role by interacting with the viral polymerase . Similar to FANCG , also XAB2 is involved in DNA repair mechanisms , in particular , in transcription-coupled DNA repair [27] . XAB2 is a host restriction factor for IAV as well as for other viruses , e . g . West Nile virus , Vaccinia virus and HIV-1 [28] . In our study , however , the overexpression of this factor neither improved nor impaired viral reproduction . In a few infected SGOs the change in various kinetic parameters should be beneficial for virus replication according to model predictions ( Fig 1 ) , e . g . a decrease in cRNA synthesis rate upon overexpression of FANCG , PLD2 or XAB2 , or an increase in the virus release rate upon overexpression of CEACAM6 , FANCG or NXF1 . Using a Monte Carlo approach , we analyzed single cell simulations using randomized SGOs parameter sets to predict virus release of MGOs . This analysis revealed that the productivity of single cells follows a log-normal distribution with highly productive cells as rare events . This finding is supported by previous single-cell analyses performed by our group , which investigated the cell-specific productivity of MDCK cells infected by IAV . In particular , they demonstrated that there is a large variability in the productivity of individual cells and that only very few cells are highly productive ( with up to 10-fold higher titers compared to the cell population average ) [29 , 30] . Furthermore , the most recent study showed that single cell virus yields are log-normally distributed [30] . While MGO simulations suggest that particular combinations of genes have the potential to yield IAV titers similar to an in silico optimized cell line with an optimal virus release rate or M1 binding rate ( open circles , Fig 7 ) , we could not generate MGOs with an elevated overall HA titer . However , it has to be taken into account that all experimental data were acquired from cell populations of genetically modified cells with different combinations and expression levels of host cell genes . Thus , beneficial host cell factor combinations in individual cell clones might be masked . More extensive screening would be required to identify and isolate individual cell clones , which reflect the features predicted in silico . The present version of the mathematical model of IAV replication is most suited to describe the impact of host cell factors that act directly on individual steps of the virus life cycle , e . g . factors that modulate the activity of the polymerases . The assumption that the influence of such factors also directly impacts kinetic parameters of the model enabled the identification of bottlenecks in virus replication that could be modulated by cell line engineering . Similar model-based approaches were performed previously by others to compare the replicative properties of different influenza virus strains [31 , 32] and virus replication with and without antiviral treatment [22 , 33] . While Binder and colleagues [34] compared low and high permissive host cells for hepatitis C virus replication that showed different intracellular basal concentrations of the same host cell factor , we applied the single cell model of IAV replication to quantify changes in key kinetic parameters of virus replication in cell lines overexpressing different host cell factors , which has not been reported before . Still , all these approaches have in common that they are solely computational , focusing on viral dynamics described by a fixed set of equations . As a result , in our study , similar ‘patterns’ of parameter changes were found for cell lines overexpressing host cell factors with very diverse functions , e . g . kImp ↓ , kVSyn→ , kCSyn→ , kMSyn↓ , kM1Bind→ and kRel ↑ for both NXF1 and CEACAM6 . Therefore , this model-based analysis can only provide indications regarding the general impact of an overexpressed host cell factor . Clearly , further in-depth characterization of the impact of host cell factors on individual steps of virus replication is required on the molecular level to fully comprehend the biological implications of parameter changes determined in the present work . To neglect details of cellular processes and pathways , e . g . cellular transcription and translation or immune response , may limit model predictions . On the contrary , the implementation of proposed functions of candidate host cell factors into the model may lead to biased interpretation of experimental data ( self-fulfilling prophecies ) . More elaborate dynamic models on virus-host cell interactions should not only account for the viral life cycle but also include a mathematical description of the cellular pathways in which the considered host cell factors are involved . Yet , the biological knowledge about how most host cell factors impact the viral life cycle is too sparse and even controversial to be readily implemented into a mathematical framework . To elucidate this in more detail can only be accomplished through experiments which analyze changes in the viral life cycle together with the dynamics of host cell factors and the activity of the corresponding cellular pathways . Regarding the further improvement of quantitative models for intracellular virus replication , this will probably be one of the most challenging tasks to be performed over the next decades . Moreover , we model viral dynamics in an average infected cell and do not account for stochastic effects that play a role at low molecule numbers , i . e . , for low MOI infections . We can therefore only estimate parameters from experimental infections performed at high MOI ( MOI ≥ 1 ) , which ensures that the majority of cells is infected simultaneously . Thus , the infection propagates synchronously in the cell population and virus release reaches steady state within 24 h . In these high MOI scenarios , replication can also be affected adversely by introducing a high number of non-infectious virions , e . g . defective interfering particles ( DIPs ) . There is already a single cell model available that also describes the impact of DIPs on virus replication [35] . However , since the intracellular mechanisms of DIP interference remain elusive , we think that , the modeling of DIP propagation in engineered cell lines seems unreasonable but should be taken into account in future studies . Usually , the significance of cellular targets identified from loss of function studies is limited , e . g . due to inefficient knockdown or off-target effects that lead to identification of false positives and false negatives ( discussed in [36–38] ) . In our study , we therefore chose host cell factors relevant for IAV replication that were not only identified in RNAi screens , but have also been described previously in separate studies , except for XAB2 . Still , the importance of these factors is mostly inferred from loss of function studies and we simply assume that if the knockdown of a host cell factor results in reduced virus growth , the overexpression of the same factor should improve virus replication . Overall , however , we found that most differences in both intracellular replication and progeny virus release were noticeable , but not statistically significant compared to parental A549 cells . Only when infected at MOI 10−4 , engineered cell lines showed higher HA titers at early time points , while the HA titers of all cell lines were similar at time of harvest ( 72 h p . i . ) . Hence , we confirmed findings of screens for which changes in virus growth were evaluated at early time points ( 12–48 h p . i . ) after infection at MOIs below one [9 , 13–15] , where a single readout is useful to identify host cell factors that have a strong impact on viral dynamics . Such factors are very interesting in the context of antiviral treatment , for which the interference with virus replication early during infection might promote viral clearance in an in vivo system . Although they are required to complete the replication cycle successfully , such factors might not even limit viral replication at their basal expression level . Hence , their overexpression would not result in any measurable changes of intracellular mechanisms . To improve vaccine production , however , the expression of host cell factors should be increased which improve the maximum cell-specific productivity . For this purpose , screening designs should be re-considered to capture not only dynamics of virus growth but also virus yield at time of harvest . Since large scale high-throughput screens are costly , a first step might be the re-evaluation of already existing screens that considered multiple time points post infection ( e . g . [10 , 39] ) . Recently , re-evaluation of primary data from various RNAi screens and different virus-host cell interaction studies , i . e . , protein-protein interactions , transcriptomic and proteomic data , revealed and validated the impact of host cell factors on virus replication , that were previously unknown [40 , 41] . This highlights the importance of study design and subsequent bioinformatical analysis , which both strongly contribute to the identification of key host cell factors for intracellular virus replication and release . Beyond that challenge , we have no indication regarding the optimal level of gene ( over ) -expression required to achieve a positive impact on virus growth , while avoiding off-target effects . In our study , we used lentiviral transduction without control of the integration site and assumed that cells , for which insertion of the overexpression constructs was beneficial , will propagate well in culture . Indeed , we saw that transduction of different host cell factors resulted in different levels of overexpression ( S3 and S4 Tables ) and surprisingly , that the cell line with a very low overexpression level of the host cell factor NXF1 was most promising with respect to early virus dynamics . In contrast , a high level of overexpression might stress the biosynthetic capacity of the cell , and result in a competition between expression of candidate genes and viral proteins . It is particularly known that the translation of viral proteins is the energetically most costly step of virus replication [42] . If the synthesis capacity of the cell is exploited by both overexpression of candidate genes and expression of viral proteins , cellular resources needed for virus growth might become limiting . Together , this might explain the observation that SGOs , in particular those showing high expression levels of the candidate gene , produce the same or only slightly higher virus yields compared to the parental A549 cell line . However , experimental proof would be needed to support these speculations . To better control overexpression levels , it might be worthwhile to explore other gene editing methods , e . g . recombinase-mediated cassette exchange [43] or CRISPR/Cas9 [44] , for target validation studies . As discussed before , some host cell factors are already enriched upon infection and it might be also interesting to follow their expression levels over time and—based on that—design an inducible overexpression system to control supply of host cell factors in a temporal manner if this is needed for their function [45 , 46] . Finally , and as shown in a first attempt in this work , mechanistic models of the virus replication cycle are indispensable for evaluation and interpretation of infection data from engineered cell lines . Thus , we envision that screening approaches focusing on virus yield at harvest time points relevant in vaccine production supported by simulation studies using mathematical models for virus replication will enable the design of novel producer cell lines with the final goal to improve cell culture-based vaccine manufacturing . In addition , the combination of both , experimental and computational , approaches using data from well-defined experimental conditions will significantly deepen our understanding of intracellular mechanisms of virus-host cell interactions and support analyses of infectious diseases and virus transmission .
The model used in this study is a detailed mathematical description of intracellular IAV replication as published previously for adherent MDCK cells [12] . It accounts for key steps of the virus life cycle , using a set of ODEs to simulate virus entry , viral RNA and protein synthesis as well as virus assembly and progeny virus release . To predict virus replication and release for A549 cells , we assumed that these do not change mechanistically , but only show differences in their dynamics due to the change in the host cell system . To capture this , we performed a re-parameterization of nuclear vRNP import , viral replication , viral transcription and virus release based on experimental data obtained for infected A549 cells ( S1 Fig ) . As an extension of the original version of this model , we also computed the percentage of nuclear vRNPs fracRnpnuc to fit measurements of nuclear vRNP import obtained by imaging flow cytometry ( Fig 3 ) . The description of the complete mathematical model can be found in S1 File . Model parameters were estimated in two subsequent steps . First , the nuclear import rate kImp was estimated by fitting the simulated fraction of nuclear vRNPs fracRnpnuc to the mean of the relative fluorescence intensity ( FI ) of the nucleus fracIntnuc determined by imaging flow cytometry ( see Imaging flow cytometry and image analysis ) . For this , we assumed that the relative increase in FI of the nucleus is correlated directly to the increase in the fraction of nuclear vRNPs caused by nuclear import of the viral genomes which can be stained by a specific antibody ( see Imaging flow cytometry and image analysis ) . In our experiments , we observed an offset for fracIntnuc of approximately 50% at the time point of infection , which is related to the background signal of the nucleus and normally comprises between 40–60% of the cell’s area evaluated during image analysis . To account for this background signal , we applied an offset to the simulation values of fracRnpnuc . Since offset values differed slightly between cell lines and showed occasionally high standard errors ( Fig 3 ) , we also estimated this offset value and optimized it with respect to the arithmetic mean and standard error of the first measurement point at zero h p . i . for each cell line . For fitting with parameter set p , we minimized the least-squares prediction error for all available data points at time point t weighted with the maximum measurement value ( Eq 4 ) . After optimization of the nuclear import rate kImp , we fitted our model to intracellular measurements of vRNA , cRNA and mRNA levels obtained from experiments at MOI 50 as well as to progeny particle numbers per cell for experiments at MOI 1 . The corresponding set of kinetic parameters p was estimated simultaneously by minimizing the least-squares prediction error based on the decadic logarithm of all state variables n , whereby the error of each variable i was weighted with its maximum measurement value ( Eq 5 ) . To synchronize infection and facilitate parameter inference , we performed infections experiments at high MOI . Thus , due to the high virus concentration at time of infection , RT-qPCR already detected vRNA copies as soon as 1 h p . i . ( Fig 4 , panel 3 ) . This value cannot be caused by an immediate uptake of all virions but rather stems from vRNAs inside virus particles and/or free vRNAs attached to the cells . Therefore , we applied the intracellular vRNA measurement value at 1 h p . i . as an offset to the simulated amount of vRNAs , as done before similarly in another modeling study of our group [22] . In contrast to this previous study , we did not apply offsets to viral mRNA and cRNA levels , as these RNA species are not part of virus particles and are usually not present in the seed virus supernatant . In particular , cRNA levels at 1 h p . i . were below or close to one copy per cell and have no significant impact on simulation results . Finally , approximately 10 copies of mRNA per cell were detected at 1 h p . i . Since mRNA synthesis starts as early as vRNPs reach the nucleus , these mRNAs are a product of primary transcription and cannot be considered as a plain mRNA offset . The parameter distributions were determined by parametric bootstrapping performing multiple model fits to 3000 random resamples from the experimental data according to their mean and standard deviation , as detailed elsewhere [47] . We set the medians of the resulting parameter distributions as parameter optima to perform simulations . For the SGO candidate FANCG , only duplicate measurements of the intracellular viral RNA were available . Therefore , we considered a relative standard error of 50% , which was the average relative standard error of all other RNA measurements performed in this study . The modeling approaches in this work are based on the simplifying assumption that each step of the virus life cycle is directly dependent on the presence of relevant host cell factors and that their influence is changed by manipulating the expression of the corresponding genes . For instance , if a host cell factor crucial for viral RNA synthesis is knocked down , the efficiency of vRNA synthesis is reduced as well , resulting in a lower vRNA synthesis rate . When the same host cell factor is overexpressed , RNA replication is enhanced , which results in a higher vRNA synthesis rate . Using this assumption , we determined the optimal value for individual kinetic parameters of the model by maximizing the number of released progeny virions at 24 h p . i . To predict biologically reasonable values , we constrained the parameter search by a lower bound of factor 0 . 2 and an upper bound of factor 5 of the original parameter values , respectively . In this study , lentiviruses were used to modify the expression of host cell factors relevant for IAV replication . Gene editing constructs delivered by lentiviruses are integrated randomly at different chromosomal locations with different transcriptional activity ( reviewed in [48] ) . Therefore , we can anticipate that individual cells within a transduced cell population will show heterogeneity with respect to levels of relative overexpression . Consequently , the transduction of more than one overexpression construct leads to an even larger heterogeneity in gene expression levels . To simulate IAV production of MGOs , we account for the non-targeted integration of multiple gene constructs by randomly compiling new parametrizations of the single cell model . More precisely , we assume that IAV can propagate in an individual cell of an MGO population with random combinations of kinetic parameters as determined before in detailed characterizations of SGO populations . In addition , to account for the adverse impacts by off-target effects , we also included the parameter set of the unmodified parental A549 cell line for randomization . To facilitate the interpretation of simulation results for MGOs , we simulated IAV replication with randomly assembled parameter sets for a single cell infection at MOI 1 for 48 h p . i . In a next step , we evaluated each simulation with respect to maximum virus yield and the time point of first virus release , i . e . , the time p . i . when the first simulated virus particle was released ( VRel≥1 ) . To assure that a sufficient number of simulations was performed that would allow reasonable conclusions on MGO single cell infections , we repeated simulations with randomized parameter sets n times until the relative deviation between the mean of n-1 and mean of n simulated maximum virus yields reached 1 x 10−8 . Model equations were solved numerically using the CVODE routine from SUNDIALS [49] on a Linux-based system . All model parameter values and initial conditions are given in S5 and S6 Tables . Model files and experimental data were handled within the Systems Biology Toolbox 2 [50] for MATLAB ( version 8 . 0 . 0 . 783 R2012b ) . Parameter values were estimated by the least-squares method as explained before ( see Parameter estimation ) , using the global stochastic optimization algorithm fSSm [51] . To determine the significance level of differences in parameter distributions between parental A549 and engineered cell lines ( SGOs ) we performed a one-sided Z test ( Gauss test ) with mean p¯ and variance σ2 taken from the empiric parameter distributions to compute the following test statistic Z: Z=pA549¯−pSGO¯σA5492n+σSGO2m ( 6 ) For this , the variance is usually normalized by the sample sizes n and m . However , we set the sample sizes to 1 instead of 3000 for the number of bootstrapped resamples , since the artificially high sample size is otherwise biasing the test result . This was also done previously by others to compare parameters of mutant to wild type viruses [31] . Following their approach , we generally assume that parameters are normally distributed . Only if parameter distributions followed a log-normal form , namely the vRNA synthesis rate kVSyn and the virus release rate kRel , the test statistics were calculated based on the decadic logarithm of these parameters . To determine statistical significance in differences of measurements from SGOs and the parental A549 cell line , the Kruskal-Wallis test was performed as available in MATLAB ( version 8 . 0 . 0 . 783 R2012b ) . Human cDNAs encoding CEACAM6 , XAB2 , FANCG , NXF1 and PLD2 were purchased from the I . M . A . G . E consortium . The cDNA sequences were amplified by PCR and cloned into the bicistronic lentiviral vector pLV-X-GFPneo . This vector was derived from pLVtTRKRAB-Red [52] by integrating the fusion gene of GFP and neomycin phosphotransferase in the second cistron . Lentiviral vectors were produced by transfecting HEK293T cells with the pLV-X-GFPneo and the lentiviral helper plasmids coding for gag-pol , Rev and VSV-G using the calcium phosphate transfection protocol as detailed in [53] . The supernatant was collected two days post transfection , filtered ( 0 . 45 μm ) , titrated and stored at -80°C . At the day of transduction , the virus supernatant was supplemented with polybrene ( 8 μg/mL ) and added to 1 x 105 A549 cells . After 6 h the virus was removed and cells were cultured in Dulbecco’s Modified Eagle Medium ( DMEM , GIBCO ) with 10% ( v/v ) fetal calf serum ( FCS , Sigma-Aldrich ) . On the day after infection , selection with neomycin was started ( 1 mg/mL G418 ) . G418-resistant cell populations were maintained as transduced populations . FACS was performed to enrich cell populations expressing eGFP . For generation of MGOs , cells were transduced with two cocktails of two to three different lentivirus stocks each on two consecutive days using MOI 1 per virus . Parental A549 cells [54 , 55] and transduced A549 cell lines were maintained in DMEM with non-essential amino acids , 10% ( v/v ) FCS at 37°C and 5% CO2 atmosphere . Prior to infection , cells were washed twice with phosphate buffered saline ( PBS ) , detached and counted using a Vi-CELL XRTM ( Beckman Coulter ) . Subsequently , 0 . 4 x 106 cells per well were seeded into multiple 12-well plates and incubated overnight . Infection was performed with an A549-adpated seed virus preparation of influenza virus A/Puerto Rico/8/34 ( #3138 , Robert Koch Institute Berlin ) which had an infectious virus titer of 1 . 08 x 108 virions per mL as determined by TCID50 ( see [56] for detailed description of the TCID50 assay ) . For infection , cells were washed twice with PBS and virus was added together with serum-free cell culture medium containing trypsin ( #T7409 , Sigma-Aldrich ) at a concentration of 1 x 10-4 units per cell . To support synchronous infection of cells , experiments were carried out at MOI 50 in a reduced volume of 300 μL per well . After 30 min , 700 μL DMEM was added to compensate for liquid losses through evaporation . To investigate the nuclear import of viral genomes , cells were treated with the translation inhibitor CHX ( Sigma Aldrich ) . For this , cells were incubated for 1 h in serum-free culture medium at a CHX concentration of 100 μg per mL . Then , infection was performed by replacing the supernatant with serum-free culture medium containing seed virus , trypsin and CHX . The amount of total virus particles in the supernatant of infected cells was determined by the hemagglutination assay as described by Kalbfuss and colleagues [57] . The virus titer measured as log10 HA units per test volume ( log10 HAU per 100 μL ) can be used to estimate the concentration of hemagglutinating particles cvirus with cvirus=cEry⋅10 ( log10HAU/100μL ) , ( 7 ) assuming that one virus particle per erythrocyte is sufficient to cause agglutination [58 , 59] , where cEry denotes the concentration of chicken erythrocytes added for hemagglutination ( 2 x 107 cells per mL ) . The number of virions released per cell was assessed by dividing the virus concentration by the maximum viable cell count obtained in each experiment . Viral and cellular RNA were purified from cells using the extraction kit ‘NucleoSpin RNA’ ( Macherey-Nagel ) according to the manufacturer’s instructions . To quantify intracellular viral RNA levels of segment 5 ( encoding viral nucleoprotein , NP ) polarity- and gene-specific tagged primers ( listed in S7 Table ) were used for reverse transcription to distinguish between the three different RNA species of the IAV genome ( as detailed in [60] ) . Reference standards were synthesized in vitro using a specific set of primers ( listed in S8 Table ) and supplemented with 350 ng of RNA from A549 cells to mimic intracellular conditions . In order to determine relative overexpression levels of host cell genes , mRNA of uninfected A549 cells was reverse transcribed using Oligo ( dT ) primers ( listed in S9 Table ) . For both , viral and cellular RNA , real time RT-pPCR was performed using the Rotor-Gene SYBR Green PCR Kit and Rotorgene Q ( Qiagen ) according to the manufacturer’s instructions . The calculation on viral RNA molecule numbers per cell was performed as described in [60] . Relative expression levels of host cell genes in SGOs and MGOs compared to the parental A549 cells were calculated by the 2−ΔΔCT method , using 18S rRNA as a calibrator [61] . For the analysis of nuclear vRNP import , 1 x 106 infected A549 cells were fixated with paraformaldehyde ( PFA ) at a final concentration of 1% ( w/v ) for 30 min on ice . Subsequently , samples were transferred to reaction tubes , cells pelleted by centrifugation ( 8 min , 300 x g , 4°C ) and resuspended in 70% ice-cold ethanol before storage at -20°C . For vRNP and DAPI staining , stored samples were centrifuged ( 8 min , 300 x g , 4°C ) and the cell pellet was resuspended in wash buffer ( PBS , 2% ( w/v ) glycine , 0 . 1% ( w/v ) bovine serum albumin ( BSA ) ) and centrifuged as before . Afterwards , the cell pellets were resuspended in 150 μL wash buffer , transferred to 96-well plates and centrifuged once more . Next , cell pellets were resuspended in 25 μL blocking buffer ( wash buffer with 1 . 1% ( w/v ) BSA ) and incubated for 30 min at 37°C . After a final washing step with 200 μL wash buffer , cells were resuspended in 25 μL antibody solution and incubated for 1 h at 37°C . The anti-NP antibody mAb61A5 that preferentially binds oligomerized NP as present in the vRNP complex , was kindly provided by Fumitaka Momose [62] . Upon incubation , cells were washed three times with wash buffer and afterwards 25 μL of Alexa Fluor 647-conjugated polyclonal goat anti-mouse antibody ( Life Technologies , #A21235 ) solution was added to the cells and incubated for 1 h at 37°C . Both the primary and secondary antibody were used at a dilution of 1:500 in wash buffer . Finally , cells were washed three times and the cell pellet was resuspended in 30 μL wash buffer with 2% ( v/v ) DAPI ( Roth , 143 μM stock solution ) for nuclear staining . After 5 min of incubation in the dark at room temperature , cells were measured using the ImageStream X Mark II ( Amnis , EMD , Millipore ) together with the INSPIRE software . For each sample 10 , 000 single cells were analyzed using the 60x magnification and the 375 nm and 642 nm lasers for excitation of DAPI and vRNP antibody , respectively . Channels 1 ( DAPI signal , CH1 ) and 5 ( Alexa Flour 647 , CH5 ) were acquired together with channel 6 ( CH6 ) , which records the bright field ( BF ) image . The laser powers were adjusted according to the value of the ‘raw max pixel’ feature that should be in the range between 200 and 1500 for single-stained positive controls . Furthermore , 1000 single positive cells were measured to adjust the compensation settings . To evaluate the localization of vRNPs only double positive single cells in focus were selected for analysis . In order to distinguish between nucleus and the whole cell , a nucleus mask and a cell mask were defined according to the DAPI signal on CH1 and the BF image on CH6 , respectively ( examples are shown in S10 Fig ) . To determine the relative fluorescence intensity of the vRNP signal ( CH5 ) located in the nucleus , the intensity of the vRNP signal within the nucleus mask was divided by the intensity of the vRNP signal within the whole cell mask . | Influenza viruses depend on cellular functions at every step of their life cycle and a comprehensive picture of virus-host cell interactions is the key to understand influenza disease and establish antiviral therapies . Over the past decade , this was supported by numerous screening approaches , which identified cellular factors relevant for intracellular virus replication . Ideally , the identification of pro-viral targets should also support the generation of cell lines to optimize influenza virus replication in cell cultures . As a first approach towards this goal , we used a mathematical model to identify mechanisms of viral growth that would be most promising targets for host cell factor manipulation . Based on predictions , we expected a significant increase in virus production if RNA synthesis and virus assembly and virus budding were perturbed , which was partially confirmed by cell lines overexpressing single and multiple selected host cell factors . However , the cell-specific productivity of engineered cell lines was not improved significantly and , according to model-based analysis , this can be explained by adverse changes in kinetic parameters of intracellular replication steps . Finally , results indicate that screening approaches should focus on late time points post infection to identify targets for engineering of cell lines that support high-yield vaccine production processes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2019 | Model-based analysis of influenza A virus replication in genetically engineered cell lines elucidates the impact of host cell factors on key kinetic parameters of virus growth |
Membrane fusion is essential to both cellular vesicle trafficking and infection by enveloped viruses . While the fusion protein assemblies that catalyze fusion are readily identifiable , the specific activities of the proteins involved and nature of the membrane changes they induce remain unknown . Here , we use many atomic-resolution simulations of vesicle fusion to examine the molecular mechanisms for fusion in detail . We employ committor analysis for these million-atom vesicle fusion simulations to identify a transition state for fusion stalk formation . In our simulations , this transition state occurs when the bulk properties of each lipid bilayer remain in a lamellar state but a few hydrophobic tails bulge into the hydrophilic interface layer and make contact to nucleate a stalk . Additional simulations of influenza fusion peptides in lipid bilayers show that the peptides promote similar local protrusion of lipid tails . Comparing these two sets of simulations , we obtain a common set of structural changes between the transition state for stalk formation and the local environment of peptides known to catalyze fusion . Our results thus suggest that the specific molecular properties of individual lipids are highly important to vesicle fusion and yield an explicit structural model that could help explain the mechanism of catalysis by fusion proteins .
Membrane fusion is critical to eukaryotic cell function; cells rely on fusion for vesicle trafficking and secretion , and viruses such as influenza and HIV utilize fusion to infect target cells . This poses a fundamental biophysical question: how do two lipid bilayers merge in a targeted manner without rupture , and how do proteins catalyze this process ? Viruses in particular are faced with a host membrane not designed to be permissive to viral entry and must alter host membrane properties to achieve fusion . Simply bringing the viral and cellular membranes together is not sufficient for physiological fusion; mutagenesis experiments in influenza [1] , [2] and parainfluenza virus [3] have demonstrated that mutations to either the viral transmembrane anchor or the fusion peptide inserted in the host membrane can block fusion . In some cases [3] , these mutations can be rescued by independently altering membrane properties , suggesting a direct connection between fusion peptides and lipid dynamics . The stalk model for membrane fusion proposes that proteins catalyze the formation of a series of lipidic fusion intermediates: the outer leaflets of each bilayer merge first , followed by opening of a fusion pore and merger of the inner leaflets [4] . There is strong indirect support for this model [4]–[8] , and stalk structures have been observed in artificial model systems [9] , but direct observation of fusion stalks in physiological membranes is extremely challenging due to their transient nature and small size . Molecular simulations provide an alternative way to study these processes and can also provide atomic detail of the fusion mechanism and transition state , yielding insight into the mechanism of biological catalysis of fusion . Vesicle fusion has previously been modeled with continuum approaches [8] , [10]–[15] or coarse-grained simulation [16]–[19] , both of which have made important contributions to refining the stalk hypothesis and outlining fusion mechanisms . One previous high-resolution simulation started from a pre-constructed stalk state , due to computational limitations , and examined a vesicle fusing to itself through a simulation boundary [20] . However , complete simulation of fusion in atomic detail has long been an important goal towards understanding atomic-level effects such as membrane dehydration and bilayer breakup upon stalk formation [21] , [22] . In cells , vesicle fusion is typically catalyzed by proteins . To understand the mechanism of this catalysis , we first wish to consider the biophysical nature of fusion , its transition state , and the surrounding molecular events . We have therefore performed atomic-resolution simulations both of complete vesicles fusing and of hemagglutinin fusion peptides interacting with lipid bilayers in order to examine the mechanism of vesicle fusion and especially stalk formation in more detail . The pathway for fusion that we observe in our simulations transits through stalk and hemifused intermediates largely as predicted by the stalk hypothesis , but we observe new high-resolution details important to understanding the transition state for stalk formation and thus how fusion proteins may catalyze the fusion process . To identify this transition state from simulations , we employ committor analysis [23]–[25] , a statistical means to evaluate the transition state ( as well as the full reaction pathway ) that has been frequently used in the protein folding literature [23] , [26] , [27] . To the best of our knowledge , this marks the first time such techniques have been applied to systems of this size and complexity , simulating the million-atom vesicle fusion reaction many times over . The transition state we identify is characterized by a hydrophobic nucleation event where lipid tails from opposite vesicles make contact within the intervening hydrophilic layer . This raises the pivotal question of how fusion proteins might accelerate this hydrophobic encounter . From additional simulations of influenza fusion peptides in bilayers , we believe that fusion catalysis may be partially explained by an increased rate of hydrophobic tail protrusion in the presence of fusion peptides .
To further probe the key structural features of the fusion stalk , we have performed committor analysis [23] to quantitatively identify a member of the transition state ensemble . We took simulation snapshots at 5-ns intervals from a fusion simulation and performed 20 simulations each 20 ns in length from each snapshot for a total of 400 20-ns simulations . Analysis of the resulting dataset yields the free-energy profile of a single fusion reaction ( Figure 2 ) . The transition state for stalk formation is identified via this committor analysis as the snapshot equally likely to form a stalk or remain as a contact patch . We confirm that the contact patch structure described above is metastable state or local free energy minimum , neither breaking apart rapidly nor rapidly proceeding to stalk formation . We find contact patches to be metastable for the tens to hundreds of nanoseconds , depending in part on the lipid composition . After formation of the contact patch but prior to the transition state , the water layer between vesicles thins substantially ( Figure 3 ) . The transition state occurs when a pair of lipid tails from opposing vesicles ( Figure 4 ) make contact in the intervening polar layer . This creates a small hydrophobic region , which either breaks apart and returns to the contact patch structure or grows to form a stable stalk . Contact patch formation and water layer thinning are quantified in Figure 5 . Analysis of additional independent fusion simulations confirms this lipid tail contact to be a consistent feature of stalk formation . At the time of contact , lipid tails bulge slightly into the hydrophilic layer and make contact to nucleate a stalk , but they are not grossly flipped , remaining roughly tangent to the vesicle surface ( Figure 6 ) . In our simulations , these bulging tails make contact in the polar layer between bilayers rather than inserting into the opposite bilayer . This is a difference from previous coarse-grained simulations [17] and may reflect the increased chain entropy of atomic-resolution lipids compared with coarse-grained simulations . These simulations suggest that the defining event for fusion occurs when two lipid tails from opposing vesicles make contact through the hydrophilic layer . To first order we can assume these bulging or protrusion events to be independent , making the nucleation probability proportional to the number of contacting lipid pairs , or equivalently the contact area A ( t ) at time t divided by the area per lipid head group ρ . Since this is a second-order reaction that depends on contact by two lipid tails , the nucleation probability varies with γ2 , where γ ( t ) is the probability of a lipid tail bulging into the hydrophilic layer . This model would explain why contact patch formation increases stalk formation rates; it also provides a new context to interpret the cooperative activity of fusion proteins and how they may interact with membranes . Engagement of multiple proteins , particularly in the ring arrangement proposed to drive fusion [39] , will help promote larger contact patch formation , thus catalyzing stalk formation . Fusion proteins have also been proposed to catalyze fusion by disordering the lipid bilayer [40] . In our formalism , this effect could manifest as a local increase in γ , the probability of lipid tail protrusion , in the vicinity of the fusion peptides . We use simulations to examine these hypotheses regarding lipid tail protrusion in closer detail . Both tail protrusion probability and lipid tail SCD order parameters in the vesicles are uncorrelated with spatial proximity to the contact interface in our simulations ( each individual correlation coefficient <0 . 2 ) , so the formation of a contact patch does not itself increase protrusion rates . The uniformity of the lipid order parameters across the vesicle surface also argues against a phase change in the contact region; the contact region remains lamellar prior to stalk formation . Average SCD values from 5–15 ns prior to stalk formation are highly similar to those in planar POPC bilayers ( Pearson correlation coefficient of 0 . 94 between values in simulated POPE vesicles and POPC bilayers measured experimentally [41] ) . In our simulations , therefore , fusion does not occur via a large-scale “disorientation” of lipid tails in closely apposed bilayers as has been previously suggested [42] . Lipid composition does affect protrusion , as POPE vesicles have significantly higher protrusion rates than POPC vesicles ( p<0 . 02 , Kolmogorov-Smirnov test ) . We also tested the ability of influenza fusion peptides to induce the tail protrusion we observe in the transition state for fusion stalk formation . Hemagglutinin fusion peptides ( HA2 residues 1–20 ) were simulated in POPC bilayers at a peptide:lipid ratio of 3∶500 for 200 ns . Such peptides have previously been studied via molecular dynamics [43]–[46] and predicted to have a disordering effect on bilayers . In our simulations , hemagglutinin significantly increased lipid tail protrusion in nearby lipids but not the bilayer as a whole ( Figure 7 ) : lipids within 5 Å of the peptides exhibited significantly increased protrusion frequencies compared to lipids greater than 20 Å away ( p<0 . 02 via Kolmogorov-Smirnov ) . No such effect was seen in membrane-inserted ion channel used as a negative control , and at >20 Å the protrusion probability was identical within error to protein-free bilayers . This local increase in protrusion does not solely account for the catalytic activity of fusion peptides , but it explains an important contribution to increased stalk formation rates . Most importantly , it provides a explicit lipid structural model for the general disordering effect that fusion peptides are thought to have on lipid bilayers to induce fusion [47] . In our simulations , fusion occurs via the following pattern: formation of a contact patch between the two vesicles precedes stalk formation . In this contact region , we observe thinning of the water layer between vesicles . Stalk formation is nucleated by a stochastic event: hydrophobic contact between a single pair of lipid tails that bulge into this water layer . If we approximate the protrusion of any single lipid tail as a statistically independent event , we derive the following model for the probability of stalk nucleation in any given time interval Δt:where A ( t ) is the contact patch area at time t , ρ is the area per lipid head group , and γ ( t ) is the probability of any single tail protruding at time t , and n is the number of contacting tails required to form the transition state . Under our simulation conditions , the transition state contains one tail from each vesicle , so in this case n = 2 . As expected for a stochastic encounter in a planar region , most stalks form off-center . This is consistent with previous simulation reports [17] and follows trivially from our model—since P ( nucleation ) is proportional to the contact area A , it varies with r2 , where r is distance to the contact patch center—but this is not how such stalks are intuitively envisioned . The vesicles simulated here are substantially smaller than either synaptic vesicles or viral particles . This was done for reasons of computational tractability , as smaller vesicles contain fewer atoms and higher membrane curvature increases the rate of fusion pore formation . Our prediction of a flattened pre-stalk intermediate in small , highly-curved vesicles is thus particularly interesting , as flattened structures would be even more favorable in larger vesicles with lower average curvature . Compared to our small simulated vesicles , we expect physiologic fusion from a comparable activated intermediate to proceed more slowly . The kinetics of fusion are difficult to separate experimentally from the generation of an activated complex; physiologic rates have been measured as fast as ∼200 µs from calcium trigger to fusion [48] while reconstituted systems are typically slower , as fast as milliseconds for synaptic fusion [49] , [50] or in the milliseconds to seconds range for viral fusion [51] , [52] . These atomic-resolution simulations suggest a structural model of the transition state that could explain many aspects of fusion protein activity . To generate this structural model , we have combined parallel simulations on traditional supercomputers with many shorter simulations in a distributed computing environment to apply committor analysis to million-atom systems . In the resulting model , membrane bending by fusion protein assemblies accelerates fusion in part by driving contact patch formation . The tail protrusion induced by influenza peptides in our simulations suggests a mechanism for fusion catalysis by bilayer disordering . Other proteins such as parainfluenza virus F protein [3] and synaptotagmin [53] that are thought to catalyze fusion in part by membrane perturbation near the site of stalk formation might also act in part by catalyzing lipid tail protrusion . This suggests the hypothesis that increased lipid tail protrusion could provide a common physical mechanism of catalysis for structurally diverse proteins: class I viral fusion peptides , membrane-associated loops of class II fusion proteins , and neuronal synaptotagmin .
Each 15-nm vesicle was composed of 877 POPC or POPE phospholipids using the Berger simulation parameters [54] . The crosslinker structure was -CO ( CH2 ) 4CO- , connected to POPC lipids via an amide linkage to the headgroup nitrogen . Individual vesicles were first equilibrated in the TIP3P explicit solvent model of water [38] . Pairs of vesicles were then placed at 1 nm separation in a hexagonal box with sides 21 nm and height 32 . 5 nm and solvated in TIP3P water with or without 150 mM NaCl , leading to a system size of over a million atoms . Simulations were run using Gromacs 4 . 0 [55] under constant temperature and pressure using Berendsen pressure coupling and the velocity-rescaling thermostat at 310 K [56] . All covalent bond lengths were constrained using LINCS [57] , and long-range electrostatics were computed every step using Particle Mesh Ewald ( PME ) [58] . The amine hydrogen atoms on POPE were converted to virtual interaction sites [55] to enable longer time steps by constraining the only polar hydrogens in the lipid system . The atomic coordinates are constructed every step , and forces acting on them are interpolated back onto the mass centers . This approach has been shown to conserve energy [59] , but we also checked the model by testing both 2 fs or 4 fs timesteps , with equivalent results for a pair of full fusion trajectories . Hemagglutinin fusion peptides were simulated based on PDB structure 1IBN [60] using the AMBER03 force field to model the amino acids [61] . 3 copies of the fusion peptide were placed in a bilayer as reported previously [44] for a total peptide:lipid ratio of 3∶500 , solvated with TIP3P water and 150 mM NaCl , and simulated for 200 ns using 2 fs timesteps , PME electrostatics , and constant pressure and temperature conditions with at 300 K with semi-isotropic pressure coupling . Lipid tail protrusion rates were significantly increased for lipids with an average distance of 5 Å to the closest peptide atom ( p<0 . 02 , Kolmogorov-Smirnov test ) . Simulations of a GLIC ion channel based on PDB structure 3EI0 [62] in a POPC bilayer were used as a negative control and showed no such increase . Tail protrusion was defined as any carbon in the lipid tail protruding more than 1 Å beyond the phosphate group . Each long fusion simulation was run on 128 cores of a cluster using Intel Clovertown or Harpertown CPU's respectively connected by an Infiniband network; the 400 “shooting” simulations were each run on 8–16 cores using the Folding@Home distributed computing network [63] . The aggregate length of vesicle fusion simulations was 10 microseconds . | Membrane fusion is a common underlying process critical to neurotransmitter release , cellular trafficking , and infection by many viruses . Proteins have been identified that catalyze fusion , and mutations to these proteins have yielded important information on how fusion occurs . However , the precise mechanism by which membrane fusion begins is the subject of active investigation . We have used atomic-resolution simulations to model the process of vesicle fusion and to identify a transition state for the formation of an initial fusion stalk . Doing so required substantial technical advances in combining high-performance simulation and distributed computing to analyze the transition state of a complex reaction in a large system . The transition state we identify in our simulations involves specific structural changes by a few lipid molecules . We also simulate fusion peptides from influenza hemagglutinin and show that they promote the same structural changes as are required for fusion in our model . We therefore hypothesize that these changes to individual lipid molecules may explain a portion of the catalytic activity of fusion proteins such as influenza hemagglutinin . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"virology/host",
"invasion",
"and",
"cell",
"entry",
"computational",
"biology/molecular",
"dynamics",
"biophysics/theory",
"and",
"simulation",
"biophysics/membrane",
"proteins",
"and",
"energy",
"transduction"
] | 2010 | Atomic-Resolution Simulations Predict a Transition State for Vesicle Fusion Defined by Contact of a Few Lipid Tails |
Gene transcription mediated by RNA polymerase II ( pol-II ) is a key step in gene expression . The dynamics of pol-II moving along the transcribed region influence the rate and timing of gene expression . In this work , we present a probabilistic model of transcription dynamics which is fitted to pol-II occupancy time course data measured using ChIP-Seq . The model can be used to estimate transcription speed and to infer the temporal pol-II activity profile at the gene promoter . Model parameters are estimated using either maximum likelihood estimation or via Bayesian inference using Markov chain Monte Carlo sampling . The Bayesian approach provides confidence intervals for parameter estimates and allows the use of priors that capture domain knowledge , e . g . the expected range of transcription speeds , based on previous experiments . The model describes the movement of pol-II down the gene body and can be used to identify the time of induction for transcriptionally engaged genes . By clustering the inferred promoter activity time profiles , we are able to determine which genes respond quickly to stimuli and group genes that share activity profiles and may therefore be co-regulated . We apply our methodology to biological data obtained using ChIP-seq to measure pol-II occupancy genome-wide when MCF-7 human breast cancer cells are treated with estradiol ( E2 ) . The transcription speeds we obtain agree with those obtained previously for smaller numbers of genes with the advantage that our approach can be applied genome-wide . We validate the biological significance of the pol-II promoter activity clusters by investigating cluster-specific transcription factor binding patterns and determining canonical pathway enrichment . We find that rapidly induced genes are enriched for both estrogen receptor alpha ( ER ) and FOXA1 binding in their proximal promoter regions .
Transcription mediated by RNA polymerase II ( pol-II ) is an essential process in the expression of protein-coding genes in eukaryotes . Transcription is dependent upon a number of sequential and dynamic events , such as recruitment of pol-II to the transcriptional start site , activation of pol-II through phosphorylation of its C-terminal domain , elongation of the nascent transcript through the transcribed region and termination [1] . Each of these steps may be rate-limiting and can therefore affect the level of gene expression . In this paper , we describe a simple probabilistic model of transcription whose parameters can be inferred using time-series data such as pol-II ChIP-Seq data [2] or nascent transcript measurement by GRO-Seq that reports markers of transcriptional activity [3] . This model can be used to identify transcriptionally engaged genes , estimate their transcription rates and infer transcriptional activity adjacent to the promoter . The transcriptional dynamics of estrogen responsive genes in a breast cancer cell line were described by fitting this model to pol-II ChIP-seq time course datasets . Chromatin immunoprecipitation , in conjunction with massively parallel sequencing ( ChIP-seq ) evaluates interactions between proteins and DNA , and , for example , can be used to monitor the presence of pol-II on DNA . Estimating the amount of pol-II associated with a transcribed gene provides a measure of transcriptional activity [2] . Sequential measurement of pol-II occupancy on genes released from transcriptional blockade , for example , in response to stimuli , reveal a wave of transcription moving through the body of the responding transcript . A number of studies have attempted to determine the rate of transcription through modelling the dynamics of pol-II . Darzacq et al . fit a mechanistic model of pol-II transcription to nascent RNA data at a single locus and obtained a transcription speed of 4 . 3 kilobases per minute [4] . Wada et al . activated transcription of genes greater than 100 kbp in length and estimated the transcription speeds using a model that measures an intronic RNA signal through taking advantage of co-transcriptional splicing . They obtain an average transcription rate of 3 . 1 kbp min−1 [5] . Singh and Padget ( 2009 ) reversibly inhibit transcription to determine the transcription rate of 9 genes , all of which were greater than 100 kbp which had an average transcription rate of 3 . 79 kbp min−1 [6] . The data used in these studies have good temporal resolution ( e . g . samples every 7 . 5 min in [5] ) and reliably allow fitting of mathematical models or the direct measurement of transcription speed , however , only for a limited set of long genes . In contrast , high throughput data sets such as ChIP-Seq , can be used to uncover transcription dynamics genome-wide but typically have much lower temporal resolution , motivating the development of alternative modelling approaches that report genome-wide transcription rates . One way around the low temporal resolution of typical high-throughput time course data is to employ a non-parametric model of the biological signals of interest . In many cases we expect these signals to vary continuously and smoothly in time , when averaged over a cell population , and a Gaussian process model provides a convenient non-parametric model in such cases [7] . Gaussian processes have recently found applications in a range of biological system models [8]–[11] . Here we present a Gaussian process model of transcription dynamics which can be fitted to genome-wide pol-II occupancy data measured using ChIP-Seq . The model describes the movement of pol-II through the gene body and combines a flexible model of promoter-proximal pol-II activity with a reliable estimate of transcription speed . By identifying genes which fit the model well , we provide a useful method to identify actively transcribed genes . The model does not assume a constant transcription speed and can therefore identify variable rates of transcription , for example due to transcriptional pausing . Model parameters are inferred using either maximum likelihood ( ML ) estimation or via Bayesian inference using Markov chain Monte Carlo ( MCMC ) sampling . The Bayesian approach provides confidence intervals for parameter estimates and can incoporate priors that capture domain knowledge , e . g . the expected range of transcription speeds , based on previous experiments . We fit our model to a pol-II ChIP-Seq time course dataset from MCF7 breast cancer cells stimulated with estradiol . The model is used to identify the set of transcriptionally engaged genes and estimate their mean transcription rate and transcriptional activity near the promoter . By clustering promoter activity profiles , potential co-regulated groups of genes are identified , particularly those that respond rapidly to estrogen signalling . Subsequent characterisation of transcription factor ( TF ) binding sites in proximity to the promoters of genes within clusters provides a means of classifying groups of promoters that are responsive to the binding of specific combinations of TFs . Additionally , publically available ChIP-Seq datasets of TF profiles from the same system were used to identify cluster-specific patterns in TF-binding . The rates of transcription estimated by our model are consistent with the literature [4] , [5] but with the advantage that our method allows the computation of transcription speeds genome-wide . Our methodology has a number of advantages . We do not require data with high temporal resolution , making it feasible to model transcriptional dynamics genome-wide using ChIP-Seq or GRO-Seq time course data . We infer transcription rates for all genes in an unbiased manner and by using Bayesian parameter estimation we are able to associate our transcription rate estimates with confidence intervals . Our model is non-parametric and therefore does not make very strong assumptions about the temporal changes in transcriptional activity . Fitting the model genome-wide allows us to identify and filter out transcripts where pol-II does not travel down the gene body . This provides a principled method to identify responsive genes , in particular , early acting estrogen responsive genes in the specific application considered here . Since our model does not enforce a uniform transcription speed over the entire gene body , we can take into account phenomena such as pol-II pausing which would result in a non-uniform transcription speed . We also use this model to infer the promoter activity of transcriptionally engaged genes , to identify co-regulated gene modules downstream of estrogen signalling .
A Gaussian process ( GP ) is a distribution over the space of functions . This distribution is completely specified by a mean function and a covariance function . A function is said to be drawn from a Gaussian process if at any finite collection of points has a multivariate Gaussian distribution with mean vector and covariance matrix specified by and , respectively . GPs provide a powerful framework for non-parametric regression [7] . If a function is assumed to be drawn from a GP with known mean and covariance function , we can infer the function value and associated uncertainty at unobserved locations given noise-corrupted observations . GPs have recently been applied in modelling biological systems , e . g . modelling protein concentrations as latent variables in differential equation models of transcriptional regulation [8] , [9] and modelling spatial gene expression [11] . Here we introduce a novel application of GPs to modelling the spatio-temporal dynamics of pol-II occupancy during transcription . Convolved GPs allow the modelling of correlations between multiple coupled data sources . In our case these data sources are the pol-II occupancy over time collected at different locations along the transcribed region of a gene . Modelling the data as a convolved process borrows information from these different data sources in estimating the model parameters and inferring the underlying signal in the data . Also , we find that convolved GPs are necessary to account for changes in the shapes of signals observed at different regions of the gene . In linear systems theory , the output of a linear time-invariant system whose impulse response is is given by the convolution of the input and , that is . If different sets of observations are believed to be related , they can be modeled as the outputs of different linear systems in response to a single input . If this input is modeled as a GP , then it will form a joint GP together with all the outputs and data from one output stream will be useful in inferring the rest [12]–[20] . In our case , incorporating the data from multiple spatially separated regions of the genes allows us to infer an underlying function that links all these regions . This proves useful as a summary of the transcription dynamics of the gene and we show that it provides useful insights into potential coregulation . A key component of our method involves the estimation of delay between time series observed at different segments of the gene . The study of time delay between related time series has received attention from a number of researchers for a long time [22] . The application areas range from signal processing to astronomy [23] . The classic approach to time delay estimation involves computing the cross-correlation between the related time series and determining the value of delay for which this function is maximised . Consider two signals and given by ( 9 ) where and are uncorrelated noise processes . The cross-correlation function is given by where denotes the expectation operator . The value of that maximises yields an estimate of the delay . When the signals are sampled at equally spaced time points with spacing between samples , the discrete time equivalent of is readily estimated . Let , the discrete cross-correlation is estimated as The delay is estimated by finding the value of for which is maximised . The corresponding delay estimate is . However , this approach does not work well when the time series are unevenly sampled as is the case in several astronomical and biological studies . A number of techniques have been developed to handle unevenly sampled time series including the discrete correlation function ( DCF ) [24] , and the more recent kernel based approaches [25] , [26] . The DCF is computed as follows , for all the time differences are binned into discrete bins of size . The DCF at is given by [24] , [25] ( 10 ) where ( 11 ) and and are the variances of the observation streams while and are observation error variances . In the kernel based approach of [25] , the underlying function of equation ( Equation 9 ) is modelled as the sum of a fixed number of kernels centered at the observation times . That is ( 12 ) where ( 13 ) The value of that minimises the estimation error is the delay estimate . Our implementation follows that presented in [25] where we assumed a fixed kernel width . This kernel width is determined by leave one out cross-validation . We used synthetic data and previously published experimental data to assess our novel method's performance . To generate the synthetic data , the underlying function of equation ( Equation 9 ) was given as a sum of Gaussian kernels . That is N was fixed at 20 and the observation interval . , and were generated at random with , and . A random delay was used to generate the observations which were corrupted by additive Gaussian noise with . To determine the effect of number of observations on the quality of inference we compute the Median Normalised Square Error ( MNSE ) of the estimated delay as a function of the number of observations for 50 random realisations of the the signals . We also investigated the effect of distorting the shape of the observed signals by introducing convolution . In real signals the restriction that the shape remains unchanged sometimes leads to poor results . The parameters of the smoothing kernel in equation ( 1 ) were generated at random with and . To assess performance of our method on a well characterised real-world dataset we obtained a dataset from Singh and Padgett [6] where the delay in appearance of pre-mRNA signal at exon-intron junctions was used to compute estimates of transcription speed for 9 genes . To generate the data , transcription was reversibly inhibited in vivo using 5 , 6-dichlorobenzimidazole 1-beta-D-ribofuranoside ( DRB ) and the pre-mRNA measured after the inhibitor was removed . As verified by the authors , the kinetics of pol-II and pre-mRNA are similar hence we expect good performance on this dataset to indicate applicability of our method to pol-II ChIP-seq data . To demonstrate an application to pol-II ChIP-Seq data , we apply our model to investigate the transcriptional response to Estrogen Receptor signalling . ChIP-seq was used to measure pol-II occupancy genome-wide when MCF-7 breast cancer cells are treated with estradiol ( E2 ) . Cells were put in estradiol free media for three days . This is defined media devoid of phenol red ( which is estrogenic ) containing 2% charcoal stripped foetal calf serum . The charcoal absorbs estradiol but not other essential serum components , such as growth factors . This results in basal levels of transcription from E2 dependent genes . The cells are then incubated with E2 containing media , which results in the stimulation of estrogen responsive genes . The measurements were taken at logarithmically spaced time points 0 , 5 , 10 , 20 , … , 320 minutes after E2 stimulation . Raw reads were mapped onto the human genome reference sequence ( NCBI_build37 ) using the Genomatix Mining Station ( software version 3 . 2 . 1 ) . The mapping software on the Mining Station is an index based mapper that uses a shortest unique subword index generated from the reference sequence to identify possible read positions . A subsequent alignment step is then used to get the highest-scoring match ( es ) according to the parameters used . We used a minimum alignment quality threshold of 92% for mapping and trimmed 2 basepairs from the ends of the reads to account for deterioration in read quality at the 3′ end . The software generates separate output files for uniquely mapped reads and reads that have multiple matches with equal score . We only used the uniquely mapped reads . On average about 66% of all reads could be mapped uniquely . The data are available from the NCBI Gene Expression Omnibus under accession number GSE44800 . Time series of pol-II occupancy over various segments of genes were computed in reads per million ( RPM ) [27] using BEDtools [28] , [29] . The genes were divided into 200 bp bins and the RPM computed for each bin . The occupancy in a particular gene segment was the mean RPM of the bins in that segment . Here , the gene is divided into five segments each representing 20% of the gene .
We first applied our methodology to synthetic data in order to compare its performance to other methods . We investigated the performance of five methods , namely cross-correlation ( Corr ) , DCF , the kernel approach of [25] ( Kern ) , a GP approach with no convolution ( GP-NoConv ) , and the convolved GP approach developed in this paper ( GP-Conv ) . Tables 1 and 2 show the MNSE for the different delay estimation methods as a function of the number of observations for synthetic data without convolution and with convolution respectively . Note that the kernel and DCF methods require an estimate of the noise variance and in this simulation study we provide the algorithms with the true value , but that would not be known in practice . We see that when no convolution is introduced , the kernel method performs well but is outperfomed by both GP methods . When convolution is introduced the kernel method appears to break down and as expected the GP-Conv outperforms the other techniques . We next applied the model to pre-mRNA data from Singh and Padgett [6] where the delay in appearance of pre-mRNA signal at exon-intron junctions was used to compute estimates of transcription speed for 9 genes . Figure 3 ( A ) shows the pre-mRNA signal for the SLC9A9 gene ( the same data shown in Figure 4d of [6] ) . The delays read from these plots were used in [6] to determine transcription speeds . Figure 3 ( B–D ) shows the fit obtained using the kernel method , GP-NoConv and GP-Conv respectively . Table 3 shows the delays read off the plots as well as values obtained using the five delay estimation algorithms for different regions of the nine genes presented in [6] . In each row the delay estimate with the lowest normalised square error is highlighted . Table 4 shows the MNSE for the five delay estimation algorithms for all the genes . We see that the convolved GP method developed in this paper outperforms the other techniques . This method has the added advantage of inferring a latent function which links all the observations and which can be used for downstream analysis . Also , when analysis is genome-wide , reading delays off individual plots is not feasible and furthermore when the sampling intervals are irregularly spaced assigning delays manually would be error prone . These results serve to justify the use of the convolved GP method introduced in this paper . We applied our method to a ChIP-Seq time-course dataset measuring pol-II occupancy genome-wide when MCF-7 cells are treated with estradiol ( E2 ) . For our initial experiment , we considered 3 , 064 genes which exhibit significant increase of pol-II occupancy between 0 and 40 minutes after E2 treatment . These genes were determined by counting the number of pol-II tags on the annotated genes in the RefSeq hg19 assembly at 0 and 40 minutes after E2 treatment and computing the ratio of these counts . We keep those genes where this quantity is greater than one standard deviation above the mean . For these 3 , 064 genes , we filtered out genes less than 1000 bp in length and computed model fits using the ChIP-seq time series data for the remaining 2623 genes . The estimation of the parameters for a given gene was performed using maximum likelihood with fixed at zero , and the values constrained to be equal . Intuitively , one would expect the values of delay to be non-decreasing . We therefore keep only those genes where this natural ordering is preserved for further analysis . We also discard genes with and since these are generally seen to be poor fits . Small values of arise when the data is best modelled as a noise process while large values model constant profiles which are not interesting in our analysis . This left us with 383 genes which we consider a conservative set of genes where there is evidence of engaged transcription and where the model parameters can be confidently estimated . To rank these genes we compared the log marginal likelihood of the model fit to that obtained if we assume independence between the segments , which is equivalent to setting the off-diagonal blocks in equation ( 8 ) to the zero matrix . Figure 2 ( A–F ) shows the inferred pol-II time profile and histogram of the samples of the delay parameters for three of the top 10 genes found to fit the model well . We note that a relatively small number of activated genes fit the model well . This is primarily because for shorter genes the pol-II occupancy quickly rises over the whole gene such that the temporal resolution of the data cannot capture the wave as it traverses the gene body . With a closer or more evenly spaced time course we would expect a good fit for a greater proportion of activated genes . Figure 4 ( A ) shows the linear regression plots using the delay samples for the TIPARP gene . Figure 4 ( B ) shows the histogram of speed samples from which we can compute the confidence interval for the speed estimate . The 95% confidence interval is indicated in Figure 4 ( B ) by the red triangle markers ( cf . Table 5 ) . Table 5 shows the average transcription speeds for the top 10 genes computed using the samples of the delay parameters . Figure 5 shows a box plot of the average transcription speeds computed using the samples of the delay parameters for these genes . The advantage of fitting each of the delay parameters independently instead of enforcing a linear relationship is that it allows us to take into account phenomena such as pol-II pausing and provides a means to filter genes where the values of estimated delay are not naturally ordered . Visual inspection of the inferred time series of the top ranked genes is consistent with a ‘transcription wave’ traversing the gene . The transcription wave is especially evident in the longer genes MYH9 and RAB10 . This motivates a closer look at long genes . Table 6 shows the average transcription speeds computed using the samples of the delay parameters for the 23 long genes found to fit the pol-II dynamics model well . Grouping these genes according to the magnitude of the median transcription speed allows us to compare our results to those presented previously . From Table 6 we see that 12 ( 52% ) of these genes have average transcription speeds between 2 and 4 kb per minute , a range that includes speeds previously reported in the literature [5] , [6] .
In this work we have presented a methodology for modelling transcription dynamics and employed it to determine the transcriptional response of breast cancer cells to estradiol . To capture the movement of pol-II down the gene body , we model the observed pol-II occupancy time profiles over different gene segments as the delayed response of linear systems to the same input . The input is assumed to be drawn from a Gaussian process which models the pol-II activity adjacent to the gene promoter . Given observations from high-throughput data such as pol-II ChIP-Seq data , we are able to infer this input function and estimate the pol-II activity at the promoter . This allows us to differentiate transcriptionally engaged pol-II from pol-II paused at the promoter and yields good estimates of transcriptional activity . In addition to estimating the transcriptional activity at the promoter , inferring the pol-II occupancy time profiles over different gene segments allows us to compute the transcription speed . We expect the delay parameters of different gene segments to be non-decreasing and this provides a natural way to determine genes that are being actively transcribed in response to E2 . Clustering the inferred promoter activity profiles allows us to investigate the nature of the response and group genes that are likely to be co-regulated . We found that the four clusters significantly enriched for both ER and FOXA1 binding within 40 kb according to public ChIP-Seq data were those that showed the earliest peak in pol-II activity at the promoter . ER and FOXA1 ChIP peaks in the neighbourhood of these genes were also more likely to be overlapping than the average for ChIP-identified binding events of these TFs genome-wide . This observation provides some support for the previously proposed role of FOXA1 as a mediator of early transcriptional response in estrogen signalling . These results also show that our method can help regulatory network inference . The inferred promoter activity profiles pinpoint the times of transcriptional activation very accurately without confounding transcriptional delays . As genes with similar inferred promoter activity profiles are likely to have similar TF binding profiles , they are likely to be co-regulated as well . The promoter profiles should therefore lead to more accurate predictions of regulator-target relationships using time-course-based methods ( e . g . [9] ) than using expression time course data . As well as modelling transcriptional speed and transcriptional activity profiles , the proposed modelling approach may have other useful applications . For example , recent research has uncovered a link between transcription dynamics and alternative splicing [41] . It is believed that aberrant splicing can cause disease and a number of studies have tried to understand the mechanisms of alternative splicing [42] . The proposed model can potentially be used to identify transcriptional pausing events , and such results could be usefully combined with inference of splice variation from RNA-Seq datasets from the same system . Also , with the increasing availability of high-throughput sequencing data exploring multiple layered views of the transcription process and its regulation , the convolved modelling approach developed here has the potential to be usefully applied to more complex coupled spatio-temporal datasets . | Cells express proteins in response to changes in their environment so as to maintain normal function . An initial step in the expression of proteins is transcription , which is mediated by RNA polymerase II ( pol-II ) . To understand changes in transcription arising due to stimuli it is useful to model the dynamics of transcription . We present a probabilistic model of pol-II transcription dynamics that can be used to compute RNA transcription speed and infer the temporal pol-II activity at the gene promoter . The inferred promoter activity profile is used to determine genes that are responding in a coordinated manner to stimuli and are therefore potentially co-regulated . Model parameters are inferred using data from high-throughput sequencing assays , such as ChIP-Seq and GRO-Seq , and can therefore be applied genome-wide in an unbiased manner . We apply the method to pol-II ChIP-Seq time course data from breast cancer cells stimulated by estradiol in order to uncover the dynamics of early response genes in this system . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biology",
"and",
"life",
"sciences",
"computer",
"and",
"information",
"sciences"
] | 2014 | Inference of RNA Polymerase II Transcription Dynamics from Chromatin Immunoprecipitation Time Course Data |
The species of the Rickettsia genus are separated into four groups: the ancestral group , typhus group , transitional group and spotted fever group . Rickettsia parkeri , a spotted fever group Rickettsia , has been reported across the American continents as infecting several tick species and is associated with a relatively mild human disease characterized by eschar formation at the tick feeding site , regional lymphadenopathy , fever , myalgia and rash . Currently , there are several mouse models that provide good approaches to study the acute lethal disease caused by Rickettsia , but these models can only be performed in an animal biosafety level 3 laboratory . We present an alternative mouse model for acute lethal rickettsial disease , using R . parkeri Atlantic Rainforest strain and C3H/HeN mice , with the advantage that this model can be studied in an animal biosafety level 2 laboratory . In the C3H/HeN mouse model , we determined that infection with 1x106 and 1x107 viable R . parkeri Atlantic Rainforest strain organisms produced dose-dependent severity , whereas infection with 1x108 viable bacteria resulted in a lethal illness . The animals became moribund on day five or six post-infection . The lethal disease was characterized by ruffled fur , erythema , labored breathing , decreased activity , and hunched posture , which began on day three post-infection ( p . i . ) and coincided with the peak bacterial loads . Significant splenomegaly ( on days three and five p . i . ) , neutrophilia ( on days three and five p . i . ) , and thrombocytopenia ( on days one , three and five p . i . ) were observed . Since R . parkeri is used at biosafety level 2 , the greatest advantage of this inbred mouse model is the ability to investigate immunity and pathogenesis of rickettsiosis with all the tools available at biosafety level 2 .
Rickettsia , a genus of Alphaproteobacteria that contains microorganisms transmitted by arthropods , is divided into four groups: the ancestral group ( e . g . , R . bellii ) , typhus group ( e . g . , Rickettsia typhi ) , transitional group ( e . g . , R . australis ) and spotted fever group ( SFG ) ( e . g . , R . conorii and R . parkeri ) [1 , 2] . Rickettsia parkeri , a member of the SFG , has been reported across the American continents as infecting several tick species including the Amblyomma maculatum complex ( A . maculatum , A . triste and A . tigrinum ) , A . ovale complex ( A . ovale and A . aureolatum ) , A . nodosum , A . parvitarsum and Dermacentor parumapertus [3 , 4] . Even though four R . parkeri strains have been described , until now only R . parkeri sensu stricto and R . parkeri Atlantic Rainforest strain ( ARF ) are considered pathogenic to humans . On the other hand , R . parkeri rickettsiosis ( disease caused by R . parkeri ) is the second most important tick-borne rickettsiosis in USA , after R . rickettsii rickettsiosis [5 , 6] . Cases have been described in Argentina and Uruguay [7 , 8] . Rickettsia parkeri ARF strain also causes eschar-associated rickettsiosis in Brazil and is associated with A . ovale ticks as the principal vector and probably with A . aureolatum ticks as an alternative vector [9–13] . The Colombian R . parkeri isolate ARF is phylogenetically the same based on the available sequences for five protein-encoding genes of the Brazilian RpARF strain , which was also isolated from A . ovale ticks ( a recognized vector in Brazil ) [10 , 14] and closely related to the Black Gap strain isolated from a Dermacentor parumapertus collected in Texas [3] . The Colombian and Brazilian isolates of RpARF also share conserved sequences in two of the three published intergenic spacers [4 , Londono et al; manuscript in preparation] . Rickettsia parkeri has been associated with a relatively mild human disease characterized by eschar formation at the tick feeding site , regional lymphadenopathy , fever , myalgia and rash [12 , 13 , 15] . Previously described mouse models provide a good approach to study the acute lethal disease produced by Rickettsia , such as R . australis infection of C57BL/6 or Balb/c mice , R . conorii infection of C3H/HeN mice and R . typhi infection of C3H/HeN mice [16–18] . The first one can be utilized to study a highly invasive model of rickettsial disease , because infection involves not only endothelial cells , but also perivascular cell types such as macrophages and can utilize gene knockout mice on the C57BL/6 background [16] . The second mouse model , R . conorii infection of C3H/HeN mice , is the best available model for SFG rickettsial diseases , because the principal target cells are the vascular endothelium [17] . The third one , R . typhi infection of C3H/HeN mice , is useful to study immunity and pathogenesis in typhus group rickettsial infection [18] , but all of the aforementioned models are limited to BSL3 use only . Previous animal model publications , where R . parkeri was used as infectious agent , did not state the biosafety level used [19 , 20] , but it is well accepted that the researchers use this agent at biosafety level 2 containment [21] . The aim of this study was to characterize the pathogenicity of RpARF and describe the clinical course , outcome , pathologic lesions , and anatomic distribution of rickettsiae in this new mouse model of SFG rickettsiosis that can be performed at biosafety level 2 .
A low-passage stock of RpARF isolated from a Colombian A . ovale tick and grown in Vero cells was used for animal experiments [14] . This stock was passaged eight times in Vero cells and once in C3H/HeN mice to remove Mycoplasma sp . contamination . Rickettsial identity and Mycoplasma sp . removal were confirmed by full genome MinION nanopore and Illumina sequencing . The infected Vero cells were harvested and then frozen in sucrose-phosphate-glutamate buffer solution ( 218 mM sucrose , 3 . 76 mM KH2PO4 , 7 . 2 mM K2HPO4 , and 3 . 9 mM glutamate ) . Rickettsial quantification was performed using a qPCR assay to quantify the viable bacteria . Briefly , a six-well plate with confluent Vero cells was inoculated with 1 ml of RpARF stock at a 1:100 dilution in MEM with 1% bovine calf serum ( BCS ) . The medium was aspirated from each well , and the wells were infected in triplicate or left as negative controls . The plate was centrifuged at 730 RCF for five minutes and incubated at 37°C for one hour . After one hour incubation to allow for rickettsial adsorption and entry into the cells , the plate was washed three times with warm ( 37°C ) sterile PBS to remove extracellular rickettsiae and incubated again for one hour with 1 ml of 1% BCS growth medium with DNase ( 10U/μl ) to degrade DNA of non-viable rickettsiae . The plate was washed three times followed by DNA extraction using the DNeasy Blood and Tissue Kit ( Qiagen , Hercules , CA ) . For this purpose , 200 μl of PBS and 200 μl of the buffer AL were added to each well and incubated for 2 minutes . The contents of each well were collected in a clean tube and incubated with 20 μl of proteinase K at 56°C for 10 minutes . A positive control consisted of DNA from 10 μl of the original stock with 190 μl of PBS that was extracted in parallel . The DNA was further extracted following the manufacturer’s protocol . Primers for the single copy gltA gene , CS-5 ( GAGAGAAAATTATATCCAAATGTTGAT ) and CS-6 ( AGGGTCTTCGTGCATTTCTT ) , were used for quantitative real-time PCR ( qPCR ) [22] with iTaq Universal SYBR Green Supermix ( Bio-Rad , Hercules , CA ) , and the standard curve was prepared with dilutions ( 109 to 101 copies/μl ) of a R . conorii gltA PCR fragment-containing plasmid , adapted from the assay used for Orientia [23] . The concentration of the RpARF stock used for this study was determined to be 3 . 72 x 109 viable bacteria per ml . This value was utilized to calculate the doses of rickettsiae ( 1 x 108 , 1 x 107 or 1 x 106 bacteria ) in a total volume of 200 μl . Eight week old male C3H/HeN mice from Charles River Laboratories , Inc . ( Houston , TX ) were used in this study . Since the RpARF was originally isolated at BSL-3 , by the requirements of the Institutional Biosafety Committee the experiments were performed in an animal biosafety level 3 facility , under specific pathogen-free conditions . All animal work was approved by the Institutional Animal Care and Use Committee ( protocol # 9007082 ) of the University of Texas Medical Branch-Galveston , and mice were used according to the guidelines in the Guide for the Care and Use of Laboratory Animals and comply with the USDA Animal Welfare Act ( Public Law 89–544 ) , the Health Research Extension Act of 1985 ( Public Law 99–158 ) , the Public Health Service Policy on Humane Care and Use of Laboratory Animals , and the NAS Guide for the Care and Use of Laboratory Animals ( ISBN-13 ) . To characterize the pathogenicity of our isolate of RpARF in a murine model , a dose range-finding experiment was undertaken . Groups of mice ( n = 3/group ) were inoculated intravenously ( I . V . ) with 1x106 , 1x107 or 1x108 viable bacteria or PBS ( 200 μl ) . Animals were monitored daily for signs of illness , and rectal temperature and body weight were measured until moribund or the survivors were sacrificed at day 14 post infection . To characterize the kinetics of the lethal murine model of infection with RpARF , three groups of animals ( n = 6/group ) were infected with 200 μl of the highest dose ( 1x108 ) I . V . , and one group was utilized as controls ( 200 μl of PBS I . V . ) . Due to the kinetics of the high dose infection , which resulted in a lethal moribund state by day six p . i . , infected mice were sacrificed on days one , three and five post-infection ( p . i . ) , and the animals from the control group ( n = 6 ) were sacrificed on day six . As in the first study , the animals were monitored daily . The spleen , kidneys , liver , lungs , heart and brain were collected from each animal to determine bacterial loads and evaluate histopathologic changes . We collected blood samples in BD microtainer tubes with and without K2EDTA ( Becton Dickinson , Franklin Lakes , NJ ) to analyze the blood cell counts , determine bacterial loads , and assess the serologic response . Blood samples were collected in two microtainer tubes , with and without K2EDTA , from each animal at the time of sacrifice . The complete blood cell counts were measured with a HemaVet 950FS apparatus ( Drew Scientific , Miami Lakes , FL ) . Indirect immunofluorescence assay ( IFA ) was performed to measure IgG and IgM antibodies to RpARF using acetone-permeabilized RpARF-infected Vero cell-coated slides . The slides were immersed in phosphate buffered saline ( PBS ) for 10 minutes at room temperature , transferred to blocking solution ( PBS with 1% bovine serum albumin [BSA] and 0 . 01% sodium azide ) , and incubated for 15 minutes . Sera were diluted in a series of two-fold dilutions starting at 1:64 in a solution of PBS with 1% BSA and 0 . 1% Tween-20 . Experimental samples of serum as well as one positive control and one negative control serum per slide were added to the wells and incubated at 37°C for 30 minutes in a humidified chamber . For IgM titer determination , IgG binding was blocked by treating the serum with IgM Pretreatment Diluent ( Focus Diagnostics , California , USA ) prior to assaying . The slides were washed two times for 10 minutes in PBS containing 0 . 1% Tween-20 . Secondary antibody , DyLight 488-conjugated anti-mouse IgG ( 1:15 , 000 dilution , Jackson Immunoresearch , West Grove , PA ) or FITC-conjugated anti-mouse IgM antibody , mu-chain specific ( 1:500 dilution , Vector Laboratories , Burlingame , CA ) , was added to the wells and incubated for 30 minutes in a humidified chamber . Finally , the slides were washed twice as before with the final wash containing 1% Evans blue solution , mounted with DAPI fluoromount-G ( SouthernBiotech , Birmingham , AL ) , and coverslipped . Slides were observed under a fluorescence microscope at 400X magnification ( Olympus Scientific , Waltham , MA ) . DNA was extracted from tissue and blood using a Qiagen DNeasy Blood and Tissue Kit ( Qiagen , Hercules , CA ) , following the manufacturer’s protocol . The bacterial loads were determined by qPCR using the primers CS-5 and CS-6 with the probe ( FAM—CATTGTGCCATCCAGCCTACGGT ) , and iQ Supermix ( Bio-Rad , Hercules , CA ) [22 , 24] . The standard curve was determined as described above in a qPCR assay to measure the viable bacteria . Samples were normalized using tissue weight or blood volume , and the concentration of rickettsiae is expressed as gene copies per milligram of tissue or milliliter of blood . The tissues were fixed with 10% neutral buffered formalin for two weeks and embedded in paraffin . The samples were sectioned at 5 μM thickness and stained with hematoxylin and eosin for histological analysis or processed for immunohistochemical ( IHC ) staining of rickettsial antigen [19] . The sections were incubated at 54°C overnight , deparaffinized and hydrated . After that , they were blocked with Avidin/Biotin Blocking Kit ( Life Technologies , Frederick , MD ) and treated with proteinase K ( Dako , Carpinteria , CA ) , for antigen retrieval . The sections were incubated with polyclonal rabbit anti-R . conorii antibody ( 1:300 dilution , produced in-house ) at room temperature for one hour , followed by biotinylated secondary anti-rabbit IgG ( 1:200 dilution , Vector Laboratories , Burlingame , CA ) , streptavidin-AP ( 1:200 , Vector Laboratories , Burlingame , CA ) for 30 minutes each and Fast Red ( Dako , Carpinteria , CA ) for 5 minutes . The sections were washed twice with Tris-buffered saline containing 0 . 05% Tween-20 . Slides were counterstained with hematoxylin , dehydrated , mounted with Permount and examined with an Olympus BX51 microscope ( Olympus Scientific , Waltham , MA ) . Data were analyzed using GraphPad Prism software utilizing a one-way ANOVA with Tukey’s post-test . P-values of <0 . 05 were considered to indicate statistically significant differences in each analysis .
In C3H/HeN mice , we observed that infection with 1x106 ( low dose ) and 1x107 ( mid dose ) of viable R . parkeri ARF produced dose-dependent severity of illness . The animals which received the lowest concentration doses began to show signs of illness on day three p . i . as evidenced by onset of weight-loss . Animals in the lowest dose group presented with piloerection on days three and four p . i . ; whereas piloerection on day three , and piloerection , erythema , and hunched posture were observed on day four in the mid dose group , which diminished to piloerection alone by five days p . i . By days five and six p . i . , the mice of the low- and mid-dose groups , respectively , did not have observable signs of illness through the experiment’s end ( day 14 ) . The control group did not present with any sign of illness during the experimental observation period . Infection with 1x108 ( high dose ) of viable bacteria resulted in a lethal illness . The animals became moribund on day six p . i . , when the mice had lost 14 . 6% of their initial body weight and developed hypothermia ( Figs 1A , 1B and 2 ) . From this point , all investigations were performed with the lethal dose . The disease was characterized by ruffled fur , erythema , labored breathing , decreased activity , and hunched back , which began on day three p . i . , and the severity of these signs intensified with time . The decreased body weight and body temperature observed in the high dose recipients in the dose-range study were replicated ( Fig 3A and 3B ) . Significant splenomegaly was observed on days three and five when compared with the control group ( Fig 3C ) . The most marked hematologic findings of the lethally infected animals were significant thrombocytopenia on days one , three and five p . i . that gradually progressed in severity each day and neutrophilia on days three and five p . i . ( Fig 3D and 3E ) . The animals infected with the low- and mid-doses developed a strong antibody response measured by IFA on day 14 p . i . The IgM reciprocal endpoint titers were between 1 , 024 and 4 , 096 for both groups and the IgG titers between 8 , 192 and 16 , 384 for the animals that received the low dose , and 16 , 384 and 32 , 768 for the animals that received the mid dose . In the animals that received the high dose , IgM antibody was detectable on day three with reciprocal endpoint titers between 256 and 512 , and between 512 and 1 , 024 on day five . In contrast , IgG antibody was present in only two of the six mice on day five at the cutoff reciprocal titer of 64 . The serologic results from all animals from the control groups in preliminary dose range experiment and infection experiments were negative . All tissues of RpARF-infected mice infected with the lethal dose contained rickettsial DNA . In the spleen , lung and liver , qPCR analysis revealed that the peak bacterial loads occurred on day three p . i . , with statistical significance , and decreased on day five p . i . ( Fig 4A , 4B and 4C ) . In the heart and kidney , the bacterial loads appeared to peak on day three , but without differences among the days of infection reaching statistical significance ( Fig 4D and 4E ) . Infection was observed in the brain on days three and five p . i . , with no statistical difference between the time points ( Fig 4F ) . In the blood , the peak bacterial load was observed on day five p . i . ( Fig 4G ) . All tissues from the control group were negative . No significant pathologic findings were detected in the tissues of mice on day 14 after recovery from low- and mid-dose of RpARF or in the uninfected control mice . Meningitis was observed in the brains of animals infected with the high dose of RpARF on days three and five p . i . ( Fig 5A and 5B ) . Pathology in the heart was characterized by mural and valvular endocarditis and perivascular interstitial cellular infiltration beginning on day three ( Fig 5E ) . The kidney showed inflammatory mononuclear cellular infiltration between the renal tubules and intertubular capillaries ( Fig 5G ) . Interstitial pneumonia was observed beginning on day three p . i . ( Fig 5I ) . Cellular infiltration in liver was characterized by a high ratio of polymorphonuclear cells to mononuclear cells on day three p . i . , and the inverse was observed on day five p . i . , fewer polymorphonuclear cells and an increase in mononuclear cells ( Fig 6A and 6B ) . RpARF organisms were associated with vessels of the microcirculation in areas of pathological damage in the brain , heart , kidney and lung by immunohistochemical staining ( Fig 5C , 5D , 5F , 5H and 5J ) . In the liver , RpARF was observed in hepatocytes and mononuclear cells in addition to endothelial cells ( Fig 6C , 6D , 6E and 6F ) .
To decrease the impact that infectious agents have in public health , valid animal models of infectious diseases are necessary to understand the pathophysiology of the disease and to evaluate vaccine candidates and treatments . Infection of C3H/HeN mice with low- and mid-dose RpARF resulted in dose-dependent severity , whereas infection with the high dose produced a lethal illness . The animals became moribund by day five or six p . i . The lethal disease was characterized by ruffled fur , erythema , labored breathing , decreased activity , and hunched back , which began on day three p . i . and coincided with the peak bacterial loads in some tissues ( Fig 4 ) . Other significant observations included splenomegaly ( on days three and five p . i . ) , thrombocytopenia ( on days one , three and five p . i . ) , and neutrophilia ( on days three and five p . i . ) ( Fig 3 ) , which are common findings in murine models of rickettsial infection [25] . Two previous studies established that R . parkeri causes dose-dependent severity in mice . In one study , the authors infected C3H/HeN mice with 5 . 5 x 106 of the R . parkeri Portsmouth strain , and the animals sacrificed on day seven did not develop any clinical signs of illness . The mice displayed mild to moderate splenomegaly , and R . parkeri DNA was detected in heart and lung of only 50% of the animals [26] . In the other study , 1 . 2 x 107 R . parkeri Maculatum 20 strain caused lethargy , mild hunched posture , ruffled fur and decreased activity with onset around days 6–7 p . i . These clinical manifestations persisted for 2–3 days . Bacterial DNA was detected in lung , spleen , liver and brain on day six p . i . [19] . The data from these two reports of C3H/HeN mice infected by intravenous inoculation of R . parkeri are similar to our low- and mid-dose infections , respectively . But neither of those previous works focused on the development of a lethal model of infection . The focus of one study was the development of a natural mode of transmission using ticks [19] . The objective of the second study was to develop a non-lethal mouse model to study long term progression and the subsequent recovery from SFG rickettsioses [26] . One of the benefits of our model is that it allows for the evaluation of vaccines , therapeutics and immunity against lethal SFG rickettsioses such as Rocky Mountain spotted fever . It is for that reason that the previous studies were only compared with the pilot assay ( preliminary dose range experiment ) . The high dose provides a model that is unique compared with other animal models of R . parkeri infection . A study of infection of guinea pigs ( Cavia porcellus ) , where the authors used a similar high dose , by intraperitoneal inoculation of 1 . 9 x 108 R . parkeri Black Gap strain resulted in a nonlethal illness , manifested in most animals by mild fever and mild to moderate swelling and erythema of the scrotum [3] . This strain has a close genetic identity with R . parkeri Atlantic Rainforest strain [3 , 4] . Furthermore , subclinical infection was observed in an additional study that utilized the natural mode of infection of guinea pigs by tick transmission , using A . ovale nymphs infected with the Brazilian Atlantic Rainforest strain [20] . Various efforts have been undertaken to establish appropriate animal models for Rickettsia spp . In 1908 , Howard T . Ricketts and L . Gomez described that guinea pigs could be utilized to isolate the agent and to study immunity [27] . Sammons et al . in 1977 demonstrated that infection with Rickettsia spp . was lethal in some mouse strains such as Mai: ( S ) and BALB/cJ infected with 1 . 3 x 106 plaque forming units ( pfu ) of R . akari , Mai: ( S ) infected with 1 x 106 pfu of R . sibirica , BALB/cJ infected with 1 x 106 pfu of R . australis and guinea pig infected with 1 x 107 pfu of R . rickettsii [28] . Later , in 1984 the model of C3H/HeJ mice infected with 1 x 1010 pfu of R . conorii was described as “an excellent animal model for studying the pathogenesis of R . conorii infection and for testing the immunogenic potential of experimental rickettsial vaccines” [29] . All these animal models employed intraperitoneal or intradermal inoculation as an infection route and some studies such as a R . conorii model , utilized a high concentration of bacterial inoculum [27–29] . Currently there are three valid lethal mouse models for rickettsiae that have been characterized , in which the infectious route is intravenous , C57BL/6 or Balb/c mice infected with 2 x 106 pfu of R . australis , C3H/HeN mice infected with 2 . 25 x 105 pfu of R . conorii , and C3H/HeN mice infected with 3 x 105 pfu R . typhi . All three of these models must be performed in an ABSL-3 laboratory [16–18] . Since this RpARF was isolated in a biosafety level 3 laboratory , all experiments in this work were performed in an animal biosafety level 3 ( ABSL-3 ) laboratory owing to local Institution Biosafety Committee ( IBC ) restrictions . However , when considering the severity of disease caused by R . parkeri , it should be noted that it produces a disease less severe than that caused by Ehrlichia species , which have been reclassified from BSL3 to BSL2 . There is also a precedent for conducting research on Rickettsia species at BSL2 [30] . Several low to avirulent species ( R . montanensis , R . rhipicephali , R . bellii and R . canadenesis ) can be manipulated at BSL2 . It is even recommended that the containment levels of newly discovered rickettsial species should be evaluated on a case by case basis [21 , 30] . Since preparation of this manuscript we have successfully petitioned our institution to work with this agent in the future at BSL2 . Currently , this agent is available in our BSL2 laboratory with the advantage that it can be used in institutions that only have BSL2 laboratories . Since lethal rickettsial animal models only can be performed at BSL3 , this model offers an excellent opportunity to work with fewer equipment restrictions , lower costs and greater safety . In conclusion , we have described an animal model using C3H/HeN mice infected by intravenous inoculation with a 1 x 108 dose of RpARF , which provides an opportunity to study acute lethal disease produced by SFG rickettsiae characterized by infection of endothelial cells in the brain , heart , lung and kidney , and endothelial cells , hepatocytes and mononuclear cells in liver . It is our belief that this experimental infection provides an alternative animal model in mice of rickettsial disease with the advantage that this model can be studied in an animal biosafety level 2 ( ABSL-2 ) laboratory . The benefit of using this at BSL2 is that it will allow for those scientists whose institutions do not have BSL3/ABL3 facilities to be able to conduct research on an actual lethal model of a rickettsial disease and for research equipment that is not available in BSL3 to be utilized . | Rickettsia is a bacterial genus that contains distinct species that are transmitted by arthropods . Many of these agents produce infection and disease in humans . The illness can range from very severe , such as Rocky Mountain spotted fever caused by Rickettsia rickettsii to mild human disease characterized by eschar formation at the tick feeding site and less severe symptoms caused by Rickettsia parkeri and often apparently asymptomatic seroconversion as observed with R . amblyommatis . To study these diseases , animal models are invaluable , and mouse models offer the best advantages for studies of immunity and pathogenesis because of the availability of immunologic reagents and gene knockout animals . Several mouse models are available for the study of the acute lethal disease produced by these bacteria , providing the opportunity to test different treatments and vaccine candidates . However , work with these models requires an animal biosafety level 3 laboratory . In this report , we present an alternative mouse model with R . parkeri Atlantic Rainforest strain available for investigation in a biosafety level 2 laboratory to study an acute dose-dependent lethal spotted fever group rickettsial disease with the advantage that experiments can be performed at this biosafety level . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"bacteriology",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"ixodes",
"pathology",
"and",
"laboratory",
"medicine",
"animal",
"models",
"of",
"disease",
"pathogens",
"microbiology",
"rickettsia",
"animals",
"animal",
"models",
"model",
"organisms",
"experi... | 2019 | A biosafety level-2 dose-dependent lethal mouse model of spotted fever rickettsiosis: Rickettsia parkeri Atlantic Rainforest strain |
Many genes can play a role in multiple biological processes or molecular functions . Identifying multifunctional genes at the genome-wide level and studying their properties can shed light upon the complexity of molecular events that underpin cellular functioning , thereby leading to a better understanding of the functional landscape of the cell . However , to date , genome-wide analysis of multifunctional genes ( and the proteins they encode ) has been limited . Here we introduce a computational approach that uses known functional annotations to extract genes playing a role in at least two distinct biological processes . We leverage functional genomics data sets for three organisms—H . sapiens , D . melanogaster , and S . cerevisiae—and show that , as compared to other annotated genes , genes involved in multiple biological processes possess distinct physicochemical properties , are more broadly expressed , tend to be more central in protein interaction networks , tend to be more evolutionarily conserved , and are more likely to be essential . We also find that multifunctional genes are significantly more likely to be involved in human disorders . These same features also hold when multifunctionality is defined with respect to molecular functions instead of biological processes . Our analysis uncovers key features about multifunctional genes , and is a step towards a better genome-wide understanding of gene multifunctionality .
Multifunctionality can be defined as the involvement of a gene in multiple cellular processes [1] . This can come about either because a protein coded by a gene is capable of performing distinct molecular functions [2–6] , or as a result of a single molecular function being performed in different contexts [7 , 8] . For example , pioneering experimental work led to the surprising finding that crystallins—the proteins responsible for the optical properties of the eye lens—can also play non-refractive roles and have enzymatic activity in other tissues [2] . This evolutionary strategy was named “gene sharing” [9] . Further examples of proteins performing multiple molecular functions were subsequently described: a uracil-DNA glycosylase that can also function as a glyceraldehyde-3-phosphate dehydrogenase , or the enzyme thrombin that can moonlight as a ligand for surface receptors [3] . More recently , a large-scale screening of mutants in yeast was performed to measure the pleiotropic effects of genes under different conditions [10] . In the case of pleiotropy , a gene may perform only one molecular function , but it can be involved in multiple biological processes , and its perturbation can therefore have pleiotropic consequences . Though multifunctionality has been characterized in detail only for a few case studies , it is likely to be a common phenomenon . Nevertheless , multifunctionality remains poorly understood . Fortunately , the current state of known gene functional annotations for several organisms gives us an opportunity to systematically identify multifunctional genes and analyze their properties . Earlier computational studies have attempted to identify multifunctional genes from functional annotations available for genes in different organisms . Several previous works measured multifunctionality by simply counting the number of distinct Gene Ontology ( GO ) biological process terms annotating a gene product [11–14] . While intuitive and straightforward , these approaches do not always guarantee that a gene annotated with more than one GO term is indeed involved in two distinct biological processes . In particular , this assumption is incorrect when one term is a descendant of another term in the GO hierarchy . To better handle the hierarchical organization of GO , an alternate approach considered the total number of distinct GO “leaf” terms annotating a gene [15] , and a recent analysis used semantic similarity between GO terms to identify moonlighting proteins [16] . However , problems may also arise even when two terms are in completely different branches of the ontology , as idiosyncrasies in GO may lead to similar processes being categorized in distinct places in the ontology . Methods to overcome this redundancy by focusing on a manually curated subset of terms ( e . g . , GO Slim or other gold standards [17–19] ) , even though suitable for tasks such as function prediction , can introduce a bias from manual curation to the analysis of gene multifunctionality , and also may not be generalizable as more annotations become available . Other approaches have used protein-protein interaction data and defined proteins as multifunctional if they are located at the intersection of overlapping clusters [20] . However , computationally derived clusters can differ substantially depending on the algorithm used [21] , thereby leading to imprecise views of multifunctionality . Further , using interaction data to define multifunctional genes has the obvious drawback of preventing an unbiased analysis of these genes’ network properties . In our work , we develop a computational approach to identify multifunctional genes that leverages GO functional annotations in a systematic and robust manner . To handle similar terms that appear in distant places in GO , we explicitly select sets of terms that do not co-annotate an enriched number of genes; these terms are then used to identify multifunctional genes . We apply our procedure to detect multifunctional genes to three organisms—human , fly and yeast—and then compare in each organism the properties of multifunctional genes ( and the proteins they encode ) with those of other annotated genes . Our results across these species consistently show that , as compared to other genes , multifunctional genes possess distinct physicochemical properties , are more broadly expressed across cell types and tissues , tend to be more evolutionarily conserved , are more likely to be essential , and are topologically distinct in protein-protein interaction networks , in regulatory transcription factor–gene networks and in genetic interaction networks . We also find that multifunctional genes are significantly more likely to be involved in human disorders than other genes . Overall , our analysis leads to a more complete understanding of the role multifunctional genes play in the functional organization of the cell .
We use functional annotations of genes in three organisms , H . sapiens , D . melanogaster , and S . cerevisiae , to identify multifunctional genes in each of them at the genome-wide level . To accomplish this , we use Biological Process GO annotations [22] , though in subsequent analyses we also consider multifunctionality with respect to Molecular Function . In the remaining text , when we refer to GO annotations , we refer to Biological Process terms unless otherwise specified . Our method for detecting multifunctional genes is shown schematically in Fig 1 and is briefly described below ( see Materials and Methods for details ) . The Biological Process GO is a hierarchy of terms representing different aspects of biological processes , where the terms range from very general to very specific and a relationship between terms indicates if one term implies another . We therefore start by selecting a subset of comparable terms that do not have ancestor or descendant relationships amongst themselves . This set of terms can be chosen at different specificity levels , represented by a parameter N corresponding to the number of genes annotated by a term . Lower values of this parameter produce larger numbers of more specific terms , and higher values result in smaller numbers of more general terms ( S1 Fig ) . We consider several distinct levels of specificity and call multifunctional all genes for which we find evidence of multifunctionality at any specificity level . Once the terms have been selected at a particular specificity level , we extract all genes annotated with at least two such terms . In order to select only pairs of distinct terms and make sure a gene annotated by both terms is truly multifunctional , we apply several filters to pairs of terms . From the collection of all pairs of terms at a particular specificity level , we filter out those that either share a common ancestor ( other than the root ) or have a common descendant term in the GO graph , as these events indicate that the terms are semantically related . However , this is not sufficient to claim that the remaining pairs of terms are distinct . For example , the terms aerobic respiration and mitochondrial translation do not have any ancestral or descendant term in common in the GO hierarchy graph besides the most general biological process term , but often co-annotate mitochondrial ribosomal proteins and capture semantically distinct aspects of the same function . Therefore , we further remove all pairs of terms that co-annotate more genes than expected by chance ( as detected by the hypergeometric test ) . All genes co-annotated by some pair of chosen terms passing these two filters , for any set of chosen terms at each specificity level N considered , are called multifunctional . We note that , depending upon the application , our filters can be relaxed to consider more genes as multifunctional; for example , two biological processes may be considered distinct if they share a common ancestor that is sufficiently general . However , here we aim to identify genes that have the strongest evidence of multifunctionality . In what follows , we compare multifunctional genes detected in fly , human , and yeast with all other annotated genes in these organisms in order to uncover whether there are significant differences between the two groups with respect to various biological properties . The number of multifunctional genes and the total number of annotated genes for each organism is given in Table 1 , and the actual lists of identified multifunctional genes are provided as S1 File for fly , S2 File for human , and S3 File for yeast . We note that a small number of experimentally verified human , fly and yeast genes with multiple functions are known [23 , 24] , and our method is able to successfully detect a significant fraction of these genes ( see Section 1 . 1 in S1 Text ) . We start the analysis by studying some basic physicochemical properties of proteins . First , we hypothesized that multifunctional proteins may be longer than other proteins in order to accommodate more functional domains . To test this hypothesis , we compare the lengths of proteins encoded by multifunctional and other annotated genes in D . melanogaster , H . sapiens , and S . cerevisiae , and indeed find that multifunctional genes are significantly longer than other genes ( p-values 1e-39 , 1e-9 , and 8e-12 , respectively , Mann–Whitney U test ) , on average by 39% , 16% , and 15% , respectively ( Fig 2 ) . We also observe that proteins encoded by multifunctional genes have significantly higher numbers of distinct domains per protein ( p-values 2e-7 , 1e-10 , and 2e-4 , respectively ) , on average by 17% , 13% , and 8% , respectively ( Fig 2 ) ; this is consistent with the earlier finding of a small but statistically significant positive correlation between the number of GO biological process leaf terms a gene has and its number of Pfam domains [15] . However , we also note that longer proteins have more domains , so the difference in length between multifunctional and other genes can potentially explain the observed difference in the number of domains ( see Section 1 . 2 in S1 Text ) . Another mechanism that has been proposed to play a role in protein multifunctionality is the presence of intrinsically unstructured regions , which are thought to increase the structural adaptability of interaction surfaces of proteins to allow them to bind to the same or distinct partners with different effects [25] . To determine whether multifunctional proteins tend to be more disordered , we predict the fraction of disordered residues using the IUPred program [26 , 27] , and find that multifunctional genes in D . melanogaster , H . sapiens , and S . cerevisiae have a significantly higher fraction of predicted disordered residues ( p-values 6e-21 , 7e-4 , and 3e-14 , respectively ) , on average by 26% , 5% , and 31% , respectively ( Fig 2 ) . These results are in agreement with recent analyses of disordered regions in experimentally verified moonlighting proteins and a small set of computationally inferred moonlighting proteins in E . coli [16] . Overall , we find that proteins encoded by multifunctional genes are longer , have more domains and are more disordered than proteins encoded by other annotated genes . Differential gene expression is evident across tissues and cell types . A gene expressed in different contexts may have different functions depending upon how and when it is expressed . Therefore , we hypothesized that a gene associated with multiple distinct functions may be expressed in a larger number of contexts . In order to assess the relationship between gene expression and gene multifunctionality , we use genome-wide mRNA expression data and count in how many conditions , tissues or cell types each gene is expressed . For fly , we use two datasets: ( 1 ) FlyAtlas [28] , the Drosophila microarray gene expression atlas across different tissues in larva and adult , and ( 2 ) RNA-seq data from modENCODE across many different tissues and development time points , as aggregated by FlyBase [29 , 30] . For human , we use information about organism parts in which genes are expressed , obtained from Ensembl BioMart [31] . We observe that in both human and fly , multifunctional genes are expressed more broadly than other annotated genes; that is , they are expressed in a significantly larger number of tissues or organism parts ( p-values from 7e-38 to 2e-4 , Mann–Whitney U test; Fig 3A and 3B ) . A potential mechanism for gene multifunctionality is the production via alternative splicing of multiple protein isoforms with different functions . Indeed , we observe that multifunctional genes have a significantly larger number of known isoforms in fly and human ( S2 Fig ) . If different isoforms of a gene have different expression patterns , this gene may be detected as broadly expressed in genome-wide assays , which currently report expression only at the gene level , merging information about the expression of different isoforms . Indeed , we observe a significant positive correlation between the number of isoforms per gene and the number of tissues or organism parts in which it is expressed ( S1 Table ) . However , when comparing genes with an equal number of known isoforms , we still observe that multifunctional genes are expressed in larger numbers of tissues or organism parts ( although most p-values for human are above our significance threshold of 5%; S2 Fig ) . This indicates that multifunctional genes are more broadly expressed regardless of the number of isoforms . Acquiring multiple functions may constitute a special evolutionary strategy and limit gene evolutionary rates [9] . In order to study the evolutionary dynamics of gene multifunctionality at the genome-wide level and in an unbiased manner , we use evolutionary conservation scores from phastCons [32] . Scores in phastCons are computed using phylogenetic hidden Markov models of multiple sequence alignments of D . melanogaster with 14 other insect genomes , of H . sapiens with 99 other vertebrate genomes , and of S . cerevisiae with 6 other yeast species . For each nucleotide of the genome , phastCons produces a score between 0 and 1 , where higher values indicate stronger evolutionary conservation . For each gene , we average the scores of all nucleotides of each isoform of the gene , and then average over all isoforms of the gene to obtain a single value for each gene as an estimate of how evolutionarily conserved the gene is . Previously , a positive correlation between the number of biological process GO terms a protein is annotated with and its evolutionary conservation was observed for yeast [7 , 11 , 33] . In agreement with this , we find that in fly , human , and yeast , multifunctional genes are significantly more evolutionarily conserved than other annotated genes ( p-values 5e-13 , 6e-10 and 0 . 02 , respectively , Mann–Whitney U test; Fig 4 ) . Having shown that multifunctional genes tend to evolve more slowly , we next hypothesized that multifunctional genes independently detected in different organisms may be orthologous to each other . In order to test this , we compare the property of multifunctionality for orthologous proteins from different organisms . We use information about protein orthology from P-POD [34] and count how many orthologous pairs are observed where both corresponding genes are identified as multifunctional . Between fly and human , we observe 1725 orthologous pairs of genes where one gene is classified as multifunctional in fly and the other gene is classified as multifunctional in human . To assess significance , we compute the same number when randomly reshuffling multifunctional and other annotated genes from orthologous pairs in each organism , and observe on average only 845 . 1 ± 90 . 0 orthologous pairs where both genes are classified as multifunctional; thus , the actual value is 2 . 0 times higher ( empirical p-value < 1e-3 ) . For fly and yeast , we find 388 orthologous pairs between multifunctional genes ( 2 . 1 times higher than 184 . 7 ± 20 . 2 expected by chance , p < 1e-3 ) . For human and yeast , we find 576 orthologous pairs between multifunctional genes ( 2 . 2 times higher than 267 . 2 ± 32 . 6 expected by chance , p < 1e-3 ) . We conclude that the property of multifunctionality is conserved across orthologous genes of different organisms . This observation also supports the validity of our method for detecting multifunctional genes . Functional annotations of genes are in part determined by transferring information between organisms via sequence similarity , and this could potentially confound our evolutionary analysis of multifunctionality . To address this , we repeat the analysis excluding GO annotations based on sequence or structural similarity and observe the same trends ( see Section S1 . 3 in S1 Text and S14 Fig ) . Genes responsible for multiple functions may require more complex regulatory programs to differentiate functions across multiple tissues or conditions . In order to study how regulated multifunctional genes are , we use regulatory interactions from high-throughput ChIP experiments [35–38] . For each gene , we count the number of transcription factor–target interactions this gene participates in as a target . In all three organisms , we observe that multifunctional genes are regulated by a significantly larger number of transcription factors than are other annotated genes ( p-values from 3e-54 to 7e-4 , Mann–Whitney U test; Fig 5 ) . In addition to requiring more complex regulatory programs , multifunctional genes may also be associated with more complex phenotypes that require involvement with many other genes; this would be reflected in a gene’s genetic interactions . In order to compare the distribution of genetic interactions between multifunctional and other annotated genes , we use a collection of genetic interactions curated by FlyBase [30] for fly and by BioGRID [39] for yeast . Previously , a positive correlation between the number of biological process GO annotations a gene has and its number of genetic interactions was observed for yeast [19] . In agreement with this , we observe that in fly and yeast , the number of genetic interactions is significantly higher for multifunctional genes than for all other annotated genes ( p-values 5e-25 and 2e-55 , respectively; Fig 5 ) . Moreover , in a more refined comparison for yeast , we observe that both the number of positive and the number of negative genetic interactions are significantly larger for multifunctional genes than for other annotated genes ( p-values 9e-30 and 1e-40 , respectively; Fig 5 ) . A gene associated with multiple functions may be more important for the normal functioning of the cell and therefore may potentially be more critical for survival than a gene associated with a single function . In order to test this hypothesis , we consider the relationship between gene essentiality and multifunctionality . For fly , we call essential all genes with a lethal phenotype ( as curated by FlyBase [30] ) and observe that 74% of multifunctional genes are essential , whereas only 44% of other annotated genes are essential ( p < 2e-86 , hypergeometric test; Fig 6A ) . In addition , we use data from genome-wide RNAi screens in cell lines [40] and observe that , even though only a small fraction of genes in the study overall are detected as essential , multifunctional genes have a significantly higher fraction of essential genes than other annotated genes do ( 3 . 8% and 2 . 9% , respectively , p < 0 . 046; Fig 6B ) . For human , we call essential all genes that have a mouse ortholog with a lethal phenotype ( according to MGI [41] ) . We find that 53% of multifunctional genes are essential , whereas only 42% of other genes are ( p < 7e-16; Fig 6C ) . Using data from a genome-wide RNAi screen in human mammary cells [42] , we also observe that multifunctional genes are essential significantly more often ( p < 1e-34; Fig 6D ) . In a more detailed analysis using quantitative data about essentiality in 72 human cancer cell lines [43 , 44] , we confirm that in all 72 cell lines , multifunctional genes tend to be more essential ( S3 Fig ) . Gene essentiality has been found to correlate with evolutionary rate [45 , 46] , and we observe that multifunctional genes tend to be more evolutionarily conserved; thus , the increased evolutionary conservation of multifunctional genes could potentially explain their preferential essentiality . We confirm that whether a gene is essential is correlated with its evolutionary conservation , but observe that multifunctional genes are still significantly more essential when controlling for evolutionary conservation ( see Section 1 . 4 in S1 Text ) . We note , however , that the relationship between multifunctionality and evolutionary conservation becomes much weaker when controlling for essentiality , and thus the tendency of essential genes to be more evolutionarily conserved may indeed explain the tendency of multifunctional genes to be more evolutionarily conserved ( see Section 1 . 4 in S1 Text ) . In contrast to fly and human , for yeast , when using information about essentiality for growth in rich medium , we do not observe a significant difference in essentiality: 24% of multifunctional genes and 26% of other annotated genes are essential ( p = 0 . 11; Fig 6E ) . However , in a genome-wide screen of yeast homozygous and heterozygous deletion strains across a variety of conditions , up to 97% of yeast genes are reported as essential in at least one condition [47] . Using these data , we count the number of conditions in which each gene is detected as essential , and find that multifunctional genes are essential in a significantly larger number of conditions than are other annotated genes ( p-values 2e-04 and 3e-03 for homozygous and heterozygous screens , respectively; Fig 6F ) . As multifunctional genes are more critical than other genes for the survival and normal functioning of the cell , they may potentially also be more likely to be associated with diseases . To address the relationship between gene multifunctionality and involvement in human disorders , we use the gene-disease “morbid map” from the Online Mendelian Inheritance in Man ( OMIM ) catalog [48] , and calculate the fraction of genes with an OMIM annotation among multifunctional genes found for human . We find that 32% of all multifunctional genes are involved in at least one Mendelian disorder , whereas the fraction of other annotated genes involved in at least one Mendelian disorder is 21% ( p < 8e-30 , hypergeometric test; Fig 7A ) . To further investigate the relationship between multifunctional genes and their involvement in human disorders , we look at genes involved in multiple distinct disorders . We map OMIM terms onto the Disease Ontology [49] and identify genes with at least one pair of disjoint OMIM terms ( i . e . , diseases that fall into separate branches of the Disease Ontology ) . We consider these genes to be involved in two or more distinct diseases . When considering genes involved in at least one disease from the Disease Ontology , we find that 18% of multifunctional genes are involved in at least two diseases , while only 8% of other such genes are involved in at least two diseases ( p < 4e-8; Fig 7B ) . One might expect that genes involved in more disorders , as well as multifunctional genes , are more actively studied by the research community , and that this could potentially introduce a study bias affecting our observations [12] . Using the number of PubMed publications associated with a gene as a proxy for how well studied it is , we indeed confirm that multifunctional genes are more actively studied ( S4 Fig ) ; however , even when only comparing gene sets with the same number of associated publications , we observe that the fraction of genes associated with disease is higher for multifunctional genes than for other genes ( S5 Fig ) . Overall , we observe that multifunctional genes are associated with diseases significantly more often than are other annotated genes . Genes associated with multiple functions may potentially play a more central role in the global functional organization of the cell . Large-scale networks of physical protein-protein interactions provide a comprehensive view of the cellular functional landscape . In order to study how multifunctional genes are positioned in protein interaction networks , we use interaction data curated by BioGRID [39] . We use three measures of centrality: degree , betweenness centrality , and participation coefficient . Degree is the number of interactions in which a protein is involved . Betweenness centrality is the number of shortest paths passing through a node in the network , and nodes with higher betweenness are more globally central in the network . Participation coefficient shows how well a protein’s interacting partners are distributed among clusters in the network , so that proteins with low participation are mostly interacting with proteins from the same cluster , whereas proteins with high participation have their interactions spread across many clusters . We observe that with respect to all three considered measures , multifunctional genes are significantly more central than other genes ( p-values from 2e-13 to 3e-50 , Mann–Whitney U; Fig 8 ) . However , not surprisingly , degree is correlated with betweenness and participation ( S6 Fig ) , and thus the correlation between multifunctionality and degree could potentially explain the correlation with the other two more complex measures . In order to test for this , we compare multifunctional and other annotated genes with respect to their betweenness and participation when controlling for degree distribution , and still observe that multifunctional genes have significantly larger betweenness and participation ( S6 Fig and S2 Table ) . In order to show that our observations are not affected by potential study biases , we repeat the comparisons of degree , betweenness , and participation between multifunctional and other annotated genes in networks containing only interactions from high-throughput experiments ( as reported in BioGRID [39] and HINT [50] ) and observe similar results ( S7 Fig ) . Furthermore , in order to show that potential bias in the selection of baits in these high-throughput experiments does not affect our conclusions , we compare only the number of bait-to-prey interactions reported in these high-throughput experiments . In particular , we only compare multifunctional and other genes that are baits in these experiments , and observe the same trends ( S7 Fig ) . Overall , we conclude that multifunctional genes are more centrally positioned in protein interaction networks , and this suggests that they may play an intermodular role within interactomes . The main focus of our analysis thus far has been on multifunctional genes detected using the Biological Process ontology . However , the same procedure for detecting multifunctional genes can be applied to the Molecular Function ontology instead , thereby providing an orthogonal view of gene multifunctionality . For clarity , in this section we call the genes detected as multifunctional using the Biological Process ontology as BP-multifunctional and those detected as multifunctional using the Molecular Function ontology as MF-multifunctional . We identify sets of MF-multifunctional genes for each organism and observe that MF-multifunctional genes have the same distinct biological properties when compared with other annotated genes as has been reported in the previous sections for BP-multifunctional genes ( although some p-values for yeast are above our significance threshold of 5%; see S8 , S9 and S10 Figs ) . In order to see if the involvement of a gene in multiple biological processes ( BP-multifunctional ) can be explained by multiple functions of the gene at the molecular level ( MF-multifunctional ) , we directly compare the two sets of multifunctional genes derived from the two ontologies . We observe that 12% to 35% of BP-multifunctional genes are also MF-multifunctional , which constitutes a significant overlap ( p < 6e-18; S3 Table ) , while the remainder may potentially be explained by other modes of gene multifunctionality . In contrast , a gene involved in multiple molecular functions might be expected to have these molecular functions while performing different biological processes , and indeed most MF-multifunctional genes are also BP-multifunctional ( 56% to 78%; S4 Table ) . These results are consistent with previous observations made using a simpler multifunctionality definition counting leaf GO terms associated with each protein [15] . Note , however , that the total number of MF annotations is lower than the total number of BP annotations ( S3 and S4 Tables ) , and thus the total number of genes identified as MF-multifunctional is lower than the total number of genes identified as BP-multifunctional ( S4 Table ) .
Most proteins are—at least to some extent—multifunctional . Even within this context , previous experimental studies have identified proteins that perform remarkably different molecular functions [2–6] , or that affect several distinct biological processes [7 , 8 , 10] . These findings suggest the existence of a subset of genes that are endowed with a particularly high degree of functional plasticity . There is increasing evidence that the phenomenon of gene multifunctionality is actually very common; thus , studying multifunctionality at a systems level can help elucidate the functional organization of the cell . In this paper , we introduce a computational approach to systematically identify multifunctional genes using existing functional annotations , and show that multifunctional genes are characterized by distinct properties as compared to other genes . To the best of our knowledge , our work represents the largest-scale characterization of gene multifunctionality to date , with whole-genome analysis across several organisms . As compared to other studies , our approach specifically addresses some previous weaknesses in handling GO functional annotations . In several previous publications , a simple count of the number of GO terms annotating a gene was used as a proxy for gene multifunctionality [11 , 12 , 19] . However , idiosyncrasies of GO may result in similar functions or processes being categorized in distinct places in the ontology . In our approach to identify multifunctional genes , we explicitly select semantically distinct terms that co-occur less frequently than expected by chance . Special care is also taken in gauging the effects of study bias , particularly in the case of interaction network properties and disease genes; that is , multifunctional genes may appear more often in the results of various experiments and thus be more actively studied by researchers , and this could potentially introduce a study bias in our analysis . In order to avoid this , we mostly analyze high-throughput and whole-genome data sets . When looking at associations of multifunctional genes with manually curated data ( e . g . , association with diseases ) , which could potentially suffer from study bias , we directly correct for this bias . Further , we carry out inter-species comparisons , and observe similar trends across three different organisms , thereby minimizing the effects of organism-specific annotation biases . A remaining challenge in characterizing multifunctional genes at the genome-wide level is that current knowledge about gene function is far from complete; thus , new experimental information about the function of some genes could result in their reclassification as multifunctional . In the limit , we may expect that nearly all genes are , to varying degrees , multifunctional . Nevertheless , the robustness of our results—both across a diverse set of organisms with distinct functional annotations and biases as well as within a single organism when explicitly controlling for study bias—suggests that we have identified specific biological features that are associated with the degree of functional plasticity of a gene . We find that gene multifunctionality is associated with several distinct properties that have important functional consequences . In the protein interactome , multifunctional proteins have a tendency to occupy more central and intermodular regions , even after controlling for potential study bias; this suggests that multifunctional proteins connect distinct and more specialized parts of the interactome , and are critical for information flow within the cell . Consistent with their important role within the cell , we also observe that multifunctional genes are more likely to be essential and are more often found to be associated with diseases . At the expression level , multifunctional genes are more broadly expressed across different conditions or cell types than are other genes . It is therefore possible that only subsets of functions are performed by multifunctional proteins under specific conditions or in particular cell types . We also observe that the expression of multifunctional genes appears to be finely regulated , as it involves a larger number of transcription factors than expected . At the molecular level , we find that multifunctional proteins have a larger number of unique domains as compared to other proteins; this is consistent with the wider spectrum of functions that they carry out . However , consistent with previous reports [25] , we also find that multifunctional proteins have a higher degree of structural disorder . Determining which of these properties or combinations of properties represent the main mechanism underlying the functional plasticity of a gene is of great interest . It is also possible to speculate that multifunctionality may be achieved via class-specific mechanisms where certain mechanisms may be at play only for a given class of genes . As part of our analysis , we perform a cross-genomic analysis of gene multifunctionality . We find that multifunctional genes are more evolutionarily conserved than other genes; this may be due to their being under stronger evolutionary pressure as they perform multiple functions , with different functions potentially performed in different conditions . Further , orthologous genes tend to share their propensity for multifunctionality; this suggests that the multifunctionality of many genes may have an early evolutionary origin . It also further supports the validity of our method to detect multifunctional genes , as they are uncovered in each organism independently . Our method to detect genes annotated with distinct functional terms can be applied to any of the vocabularies in GO , and this allows us to look at the phenomenon of gene multifunctionality from different perspectives . We observe , not surprisingly , that most genes identified as being involved in multiple molecular functions are also identified as participating in multiple biological processes . However , we detect many genes involved in multiple biological processes for which there is no evidence of association with multiple molecular functions . While this may be partly due to the fewer number of molecular function annotations , it also suggests that these genes may perform the same molecular function while carrying out different biological processes , depending upon a spatio-temporal context . Being able to tease apart the conditions under which a specific function is performed by a gene is an important avenue for future research in functional genomics , and could even lead to the development of a context-specific GO vocabulary . In this ontology , the terms used to annotate genes could be qualified with other terms specifying the cell type , the developmental stage , or the stage in the cell-cycle in which a given function is most likely to be carried out by a gene . In conclusion , a comprehensive understanding of gene and protein function has been a major goal of computational biology since the emergence of the field . In this work , we develop a computational method for genome-wide detection of multifunctional genes using existing functional annotations . We make a number of novel observations about gene multifunctionality across several organisms , as well as confirm some previous findings ( including many cases where only anecdotal evidence existed ) . Overall , our work contributes to a better systematic understanding of the functional landscape of the proteome , and can be the basis for future work in this direction as more specific and detailed functional genomics data become available .
Gene Ontology ( GO ) [22] terms and gene association data for each organism were downloaded from http://www . geneontology . org/ on July 12 , 2013 . For the main analysis reported in the paper , we include all functional associations with evidence codes EXP ( “Inferred from Experiment” ) , IDA ( “Inferred from Direct Assay” ) , IMP ( “Inferred from Mutant Phenotype” ) , IGI ( “Inferred from Genetic Interaction” ) , IEP ( “Inferred from Expression Pattern” ) , ISS ( “Inferred from Sequence or structural Similarity” ) , ISO ( “Inferred from Sequence Orthology” ) , ISA ( “Inferred from Sequence Alignment” ) , ISM ( “Inferred from Sequence Model” ) , IGC ( “Inferred from Genomic Context” ) , IBA ( “Inferred from Biological aspect of Ancestor” ) , IC ( “Inferred by Curator” ) , TAS ( “Traceable Author Statement” ) , and NAS ( “Non-traceable Author Statement” ) . We exclude all annotations with the qualifier NOT . We also perform additional analyses restricting ourselves to GO annotations with evidence codes EXP , IDA , IMP , IEP , IC , and TAS; these results are consistent with those reported in the main body of the paper ( see S11 , S12 , S13 Figs and S6 File ) . For all GO analysis , we use code from the project goatools ( https://github . com/tanghaibao/goatools ) . We call multifunctional every gene that is annotated with at least “two sufficiently distinct functional terms of comparable specificity , ” as explained next . First , to define terms of about equal specificity , we start with the notion of informative terms used previously in the literature [51–54] , which selects for a given N all terms that annotate ≥ N genes , but whose descendant terms annotate < N genes . However , we observe that a very general term annotating many genes may have all descendant terms annotating only small numbers of genes , even if it annotates many more than N genes . For example , a fly term imaginal disc-derived wing morphogenesis ( GO:0007476 ) annotates 508 genes , but its descendant terms annotate no more than 82 genes each ( 248 genes in total ) , and it may be undesirable to call this term informative for N ≈ 100 , as it is actually a much more general term than terms that annotate approximately 100 genes . To overcome this problem , we select the set TN of all terms which annotate ≥ N genes , but < 2N genes , and whose every descendant term annotates < N genes ( this includes terms with no descendants ) . Next , from all genes annotated by terms from TN we extract the genes annotated with at least two such terms . In order to consider annotations by distinct terms only , from the collection of all pairs of selected terms { ( t1 , t2 ) : t1 , t2 ∈ TN} , we further select pairs of terms that are sufficiently distinct . First , we filter out pairs of terms that have a common descendant term , as this may be an indication of similarity between the terms . We also remove all pairs of terms that have pairwise semantic similarity larger than zero [55]; though alternate thresholds of semantic similarity could be used , here we select only pairs of terms whose least common ancestor is the root of the ontology . Finally , terms annotating similar sets of genes may correspond to similar functions , so we filter out all pairs of terms that annotate significantly overlapping sets of genes ( hypergeometric test , p < 0 . 1 ) . A gene co-annotated by some pair of selected terms from TN passing these filters is called multifunctional . In order to focus on more specific biological process terms and avoid considering less informative ( i . e . , more general ) terms annotating a lot of genes , we require that N is not greater than a certain threshold M; we choose M = 120 for the analysis in the main text . The final set of multifunctional genes is given by the union of all sets obtained for different N , where N ranges from 10 up to M , with an increment of 10 . We compare multifunctional genes with all other genes that are annotated with any selected term from TN for N between 10 and M ( with an increment of 10 ) . We show that , for all our results , the same trends are observed when varying the parameter M ( S4 File ) , the p-value threshold in co-annotation filter ( S5 File ) , and when restricting the analysis to a subset of the most reliable GO annotations ( S11 , S12 , S13 Figs and S6 File ) . Protein ortholog information was obtained from version 4 of the Princeton Protein Orthology Database ( P-POD ) [34 , 68] . Two proteins from different organisms are considered orthologous if they belong to the same family , as detected by P-POD using either OrthoMCL or MultiParanoid . For each pair of organisms , we compute how many orthologous pairs of multifunctional genes are found where one gene in a pair is from one organism and the other gene in the pair is from the other organism . To assess significance , we repeat the computation 1000 times with randomization . In each random trial , we permute the labels of multifunctional and other annotated genes within each organism , while considering only genes involved in orthologous relationships . The orthology relationship between genes of different organisms is preserved . Then we compute the average and standard deviation of the counts in random trials along with an empirical p-value of the real count with respect to the randomized counts . The degree of a vertex is the number of interactions that the corresponding protein has in the network . The betweenness centrality of a vertex v is the number of shortest paths between all pairs of vertices in the network that pass through v , with the shortest paths between two vertices s and t weighed inversely to the total number of distinct shortest paths between them . The participation coefficient [69 , 70] of a vertex with respect to a set of clusters in a network is defined as P = 1 − ∑ i ( k i k ) 2 , where the summation is over all clusters , k is the vertex degree , and ki is the number of edges going from the vertex to vertices from the cluster i . The rationale is to have P = 0 if all edges from the vertex go to a single cluster , and to have p closer to 1 if edges from the vertex are more uniformly distributed over clusters . To find clusters in the network , we use the SPICi clustering algorithm [71] with parameters optimized with a simple exhaustive search procedure to approximately maximize Newman’s modularity [72] , as described earlier [73] . For network analysis , we use the python interface to the igraph library , version 0 . 6 . 5 ( http://igraph . sourceforge . net/ ) . | Almost every aspect of cellular function depends on protein activity . In spite of being fine-tuned to carry out highly specific functions , proteins can also multitask . Experimental studies have identified genes and proteins endowed with more than one molecular function , or participating in very different biological processes . These studies suggest that the degree of functional plasticity exhibited by proteins might go well beyond a simple “one protein—one function” relationship . However , systematic studies of the properties of multifunctional genes ( and their encoded proteins ) have been limited . Here we present a computational framework to identify putative multifunctional genes , and compare their properties with those of other genes . We find that multifunctional genes are significantly different from other genes with respect to their physicochemical properties , expression profiles , and interaction properties . We also observe that multifunctional genes tend to be more conserved , and that a greater fraction of them are associated with human disorders . Taken together , these results represent a step towards a more complete understanding of the role multifunctional genes play in the functional organization of the cell . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Genome-Wide Detection and Analysis of Multifunctional Genes |
Protein–protein interactions , a key to almost any biological process , are mediated by molecular mechanisms that are not entirely clear . The study of these mechanisms often focuses on all residues at protein–protein interfaces . However , only a small subset of all interface residues is actually essential for recognition or binding . Commonly referred to as “hotspots , ” these essential residues are defined as residues that impede protein–protein interactions if mutated . While no in silico tool identifies hotspots in unbound chains , numerous prediction methods were designed to identify all the residues in a protein that are likely to be a part of protein–protein interfaces . These methods typically identify successfully only a small fraction of all interface residues . Here , we analyzed the hypothesis that the two subsets correspond ( i . e . , that in silico methods may predict few residues because they preferentially predict hotspots ) . We demonstrate that this is indeed the case and that we can therefore predict directly from the sequence of a single protein which residues are interaction hotspots ( without knowledge of the interaction partner ) . Our results suggested that most protein complexes are stabilized by similar basic principles . The ability to accurately and efficiently identify hotspots from sequence enables the annotation and analysis of protein–protein interaction hotspots in entire organisms and thus may benefit function prediction and drug development . The server for prediction is available at http://www . rostlab . org/services/isis .
Interactions of proteins are at the heart of almost every biological process . Thus , the understanding of biological mechanisms requires the knowledge of protein–protein interactions and the molecular principles that underlie them . Large-scale studies unravel networks of protein–protein interactions in cells and identify interacting pairs of proteins [1–5] . However , to fully understand these interactions , and to manipulate them , we need to identify the residues that account for binding of the proteins and stabilizing the complexes . It has been postulated that only very few of the residues in protein–protein interfaces are absolutely essential for the interaction ( in a typical 1 , 200- to 2 , 000-Å2 interface , less than 5% of interface residues contribute more than 2 kcal/mol to binding . In small interfaces , this can mean as few as one amino acid on each protein ) [6] . These residues may be instrumental in understanding the interaction and could be desired drug targets [7] . The ability to predict hotspots on a large scale may assist in identifying , analyzing , and comparing binding sites for drugs . Given a detailed 3-D structure of a complex , the residues crucial for binding are often identifiable . The Hendrickson lab , for instance , identified the most essential binding residues from their 3-D structure of HIV glycoprotein ( gp120 ) and CD4 receptor [8] . Unfortunately , 3-D structures are available for less than 1% of all known pairs of interacting proteins . In the absence of 3-D structures , the most conclusive way to probe the importance of particular residues for interaction is to experimentally mutate them , typically to alanine , and measure the effect of this substitution on the interaction [9 , 10] . Many experiments have demonstrated that most interface residues could be mutated without affecting the affinity of the protein to its partners [11 , 12] . Those few residues that , upon mutation , change the affinity are often assumed to be the most essential for the interaction and are deemed “hotspots” [6] . The limited overlap between interface residues and hotspots is demonstrated in Figure 1 , which depicts the complex of the human growth hormone and its receptor [13] . In the bound state ( Figure 1A ) , a large patch on the surface of the receptor is buried in the interface . There are 31 residues on the receptor that are in physical contact with a hormone ( Figure 1B ) . However , mutation experiments indicate that only six of these residues are energetically crucial for the interaction ( Figure 1B ) . The ways to identify hotspots have been subject to theoretical debates . It has been pointed out that given the structural and physicochemical complexity of proteins , the physicochemical features of a protein are not a simple sum of the features of its individual residues [14] . Therefore , single mutations may not always convey accurate assessments of the contribution of a residue to the interaction [15 , 16] . The theoretical validity of this argument notwithstanding , alanine scans have become the most widely used tool for identifying binding sites . While single mutations may not be tantamount to isolating the contribution of a single residue to the interaction , they are still considered a good approximation . Here , we adopt the following operational definition: if a mutation of a residue in a protein–protein interface changes the binding energy of the protein to its binding partner substantially ( ΔΔG > 2 . 5 kcal/mol ) , then this residue is a hotspot residue . To the best of our knowledge , there is currently no method that was designed to identify hotspots from sequence . However , many methods attempt to use sequence or structure to identify which residues are located in the interface between proteins [17–32] . Many of the methods that identify residues in protein–protein interfaces reach impressive levels of positive accuracy ( residues correctly predicted to be in protein–protein interfaces as a fraction of all residues predicted to be in protein–protein interfaces; often also referred to as selectivity , or precision; Equation 1 ) . However , their coverage ( residues correctly predicted in interfaces as percentage of observed interface residues; often also referred to as sensitivity , or recall; Equation 2 ) remains fairly low . In other words , although these methods attempt to identify all interface residues ( all the residues that are colored blue or red in Figure 1B ) , they capture only a small fraction of them ( e . g . , only the green residues in Figure 1C ) . We hypothesized that the reason for the low coverage of many prediction methods might be that the residues that are missed are more similar to the general population of surface residues than to the essential residues ( i . e . , they are inconsequential for the interaction ) . Therefore , a machine-learning algorithm trained on all protein–protein interface residues may learn to disregard the non-hotspot residues as noise , and identify only hotspot residues as the signal to be learned . To test this hypothesis , we applied ISIS , a prediction method developed for the prediction of all interface residues [28] , to the task of predicting only hotspots . ISIS was never trained on hotspots ( Methods ) . Instead , we trained on all interface residues found in Protein Data Bank ( PDB ) complexes ( i . e . , all interface residues were labeled “positive , ” and all other residues were labeled “negative” ) . The features on which ISIS was trained included the sequence environment of each residue ( four residues on each side ) , the evolutionary profile of all nine residues in that window , the predicted solvent accessibility of the residue and the solvent accessibility of its immediate sequence environment ( one residue on each side ) , the predicted secondary structure state of the residue and its immediate sequence environment ( one residue on each side ) , and a conservation score for each residue . Like several other methods mentioned above , ISIS predicts residues in protein–protein interfaces very accurately ( ∼90% accuracy ) . However , at this high level of accuracy , ISIS identifies fewer than 5% of the residues that were experimentally mapped to the interface . The novelty here is that we applied a generic interface-prediction method to the specific task of identifying only the residues that are crucial for stabilizing the interactions ( i . e . , the hotspots ) . The results demonstrated a surprising overlap between two principally unrelated datasets , namely on the one hand the subset of residues that was identified by experimental alanine mutations as hotspots , and on the other hand the subset of residues predicted by ISIS to be protein–protein interface residues . We obtained a large dataset of hotspots that were determined experimentally through alanine scans ( Methods ) and assessed the performance of ISIS on these hotspots . The results confirmed our hypothesis that the residues predicted by the machine-learning method are , in fact , the hotspots . Analysis of the results indicated that accurate predictions of hotspots required the combination of sequence features , evolutionary information , and predicted structural features; all this information was generated from the amino acid sequence , suggesting that the commonalities of hotspots have been imprinted clearly onto amino acid sequences in the course of evolution .
One of the most comprehensive alanine scans of all the complexes with known 3-D structures is that between the CD4 receptor and the HIV glycoprotein gp120 . This interaction involves backbone interactions , mainly on the gp120 side . However , we focused our analysis on the human CD4 receptor . Ashkenazi et al . [33] sequentially mutated many residues in the V1 domain of the CD4 receptor and studied the effect of each substitution on the binding affinity between CD4 and the HIV gp120 protein . Using a set of specific antibodies , they also assessed which mutation had no effect on the structure . They identified 25 positions within a stretch of 94 residues on CD4 that upon substitution changed the affinity of CD4 substantially , without strongly altering the conformation of the protein . Within the same 94-residue segment ( Figure 2A ) , we predicted 30 residues as interface residues; 19 of these were found experimentally to have a strong effect on binding . Of the six residues that ISIS missed , four were next to predicted interface residues . Five of the predictions that were not confirmed experimentally were residues that were not mutated in the study . Our method uses predicted structural features ( solvent accessibility and secondary structure ) . Hence , its performance depends to some extent on the accuracy of these predictions . If we have a 3-D structure of the unbound chain , we can improve accuracy and coverage by using the experimental rather than the predicted features . For example , when we used the unbound structure of CD4 as input for ISIS , we found a few additional residues that were not identified from sequence alone . The two residues that scored highest ( i . e . , about which we were most confident that they participate in binding ) were Arg59 and Phe43 . The high-resolution structure of the complex between gp120 and CD4 complex [8] revealed two residues as the most important contacts between these two proteins: Arg59 and Phe43 . For a variety of reasons , membrane proteins are a particularly popular target for alanine scans . One such alanine scan is available for the shaker voltage-gated K+ channel [34] . Within a region of 29 consecutive residues that have been scanned , eight have a significant effect on the affinity of the channel to its inhibitors agitoxin2 and charybdotoxin . We used this region as input to our method , ignoring any available structural information , and predicted 13 residues ( Figure 2B ) . Seven of the eight residues that were found experimentally were predicted by ISIS; the only residue that was missed is buried in the structure and hence is likely to affect the interaction indirectly through a conformational change . Of the six residues in our prediction that did not coincide with the residues implicated as important by the alanine scanning , five coincided with positions that were found to have significant although less dramatic effects on binding [34] . Within our set of alanine scans , almost all binding residues predicted by ISIS were found experimentally to have significant effect on binding ( Figure 2C ) . Furthermore , more than 90% of the negative predictions ( predicted not to be involved in protein–protein interactions ) were confirmed experimentally to have no effect on the energy of binding . These results were particularly surprising in light of the fact that ISIS never explicitly evaluated any energetic parameters . Using different confidence thresholds ( i . e . , picking a different point on the curve in Figure 2C ) , it is possible to increase accuracy ( true positives/all positives ) at the expense of coverage ( true positives/predicted positives ) . Note that the results for the two examples ( Figure 2A and 2B ) discussed in detail are similar to the performance of ISIS on the entire dataset of 296 mutations .
We used ISIS to represent methods that predict interface residues at high accuracy and low coverage . The results suggested that the system of neural networks that underlies ISIS learned to identify the hotspots , despite the fact that they were only a small subset of the samples that were labeled as interaction residues . The system effectively disregarded most of the residues observed in interface ( i . e . , the pupil [neural network] clearly ignored the teacher [labeled data] ) . We found that the residues ignored were mostly non-hotspot residues . These results indicated that the biophysical common denominators of hotspots are so pronounced that the neural networks could identify them without specific labeling in the training phase . What are these features that are common to hotspots ? Unfortunately , we cannot simply list a few rules or features to describe these commonalities . The neural networks identified a set of complex nonlinear correlations between the input features we used and hotspot residues . It is impossible to translate the subtle and complex dependencies that were identified by the neural networks into simple explanations , or a set of rules , in English . However , it is possible to infer which features are more or less relevant . To that end , we trained several systems using different combinations of input features . Neural networks that were trained only on the sequence environment of interface residues performed only slightly ( although significantly ) better than random ( unpublished data ) . Adding evolutionary information significantly improved performance on both interface residues and on hotspots . This result was somewhat surprising given that the conservation of predicted hotspots was only marginally different from that of all other residues ( Figure 3 ) . Conversely , predicted non-hotspot residues were only marginally less conserved than the background . In other words , although the overall difference in conservation was marginal , the addition of this information to the neural network input substantially improved performance . Apparently , the neural networks have learned to distinguish between conservation that is indicative of hotspots , and conservation that is not . Strikingly , they did so without being trained on hotspots . This underscores why linear combinations of input features did not suffice and why the extraction of singly important commonalities would at best be misleading . The analysis of the contribution of each feature suggested that successful predictions of hotspots required the combination of all features . However , even when some of these features were not available , ISIS still could provide accurate predictions ( e . g . , 15% of the proteins found less than ten homologues in today's databases ) . For these proteins , the success in predicting hotspots was lower , but still significantly higher than random ( at 70% positive accuracy , >10% of the experimentally determined hotspots were identified compared with about 70%/20% for all proteins; Figure 2 ) . We did not benchmark the ability of prediction methods other than ISIS to predict hotspots . The main reason was that no existing method ( including ISIS ) was designed to predict hotspots . The ability of ISIS to identify hotspots is an unintended consequence of the power of neural networks . Therefore , when comparing ISIS with other methods , one should remember that this comparison does not benchmark these methods in the task for which they were originally developed . Still , the question remains of whether or not any method designed to predict interface residues could predict hotspots at levels of accuracy as high as the ones we reported for ISIS . To address this question , we applied a few representative interface prediction methods to the task of predicting hotspots . In particular , we chose methods that rely on a different input feature . Analysis of the results indicated that methods that did not rely on a combination of physicochemical features , evolutionary conservation , and structural features failed to identify hotspots . We applied several prediction methods that were designed to identify interface residues to the task of predicting hotspots . To eschew obfuscation: our aim was not to benchmark methods not designed to identify hotspots . Instead , we applied these methods to narrow down the features needed to successfully predict hotspots . The evolutionary trace ( ET ) method [35] correlates evolutionary importance of residues with their importance for function . We used ET to represent the approach that relies predominantly on evolutionary conservation . Gallet et al . [22] have attempted to predict interaction sites from simple biophysical features; the method computes the hydrophobic moment [36] around each residue based on its sequence environment to determine whether this residue could be a binding site . ProMate [26] extracts its input from the 3-D structure of an unbound protein; we used it to represent methods that rely on experimentally determined 3-D structures . We also included another method that predicts interfaces exclusively using amino acid information ( and no aspects of predicted structure or evolutionary profiles ) [29] . We arbitrarily chose the operating point at which the coverage of hotspots was 15% ( Methods ) and checked the accuracy of each method for this coverage ( Figure 4 ) . ISIS and ProMate , the two methods that were most successful , use physicochemical features , evolution , and structural features . ISIS is the only sequence-based method , and the structural features it uses are based on predictions . ProMate , which relies on the 3-D structure , performed even better . The conclusion of this analysis is that no single feature suffices to characterize hotspots . Rather , it takes a complex combination of the aforementioned features that defines a residue as a hotspot . It is apparent that the neural networks identified some common denominators between hotspots that distinguish them from other interface residues . This question is hard to address given our current gold standard ( namely the dataset of experimental alanine scans ) . The number of features we use for the prediction ( 189 ) is greater than the number of positive data points in our set of alanine scans . To determine to what extent each input feature differentiates between hotspots and other interface residues , we need a substantially larger dataset of hotspots and non-hotspot residues . This could be achieved if we assume that ISIS indeed identifies hotspots . Thus , by running ISIS on a large dataset of interface residues , we can create a large dataset of predicted hotspots and a large dataset of interface residues that are predicted not to be hotspots . Then , we can use these large datasets to analyze the characteristics of hotspots versus the characteristics of other interface residues . We did this using the large dataset of interface residues that was used as a test set for training ISIS . On this dataset we compared the residues that were classified by ISIS as positive ( i . e . , hotspots ) with those that are annotated experimentally as interface residues but are classified by ISIS as negatives . Table 1 is based on the multiple sequence alignment of each protein in this dataset . For each interface residue , it shows the average occupancy of its position by each type of amino acid . We also present the average occupancy of each residue in the alignment for experimentally determined hotspots ( through alanine scan ) . These values are presented in parentheses , as the data that underlie them are sparse ( only 100 positions ) . Note that for some amino acids there are significant differences between hotspot and non-hotspot interface residues , while for others there are no substantial differences . Table 1 also presents the p-value for the difference based on a t-test . Note , for example , the 400% overrepresentation of arginine in predicted hotspots ( and the extremely low p-value ) with reference to other interface residues . However , the percentages of lysine are virtually the same for both categories . Thus , it is not simple considerations of hydrophobicity that characterize hotspots . Four aliphatic residues are depleted in hotspots ( A , V , I , and L ) , while amide side chains are overrepresented ( N and Q ) . However , the role of aromatics is unclear since tyrosine is enriched in hotspots , phenylalanine is depleted , and tryptophan has similar propensities across the interface . The experimental values ( shown in parentheses ) are very close to the values obtained for the predicted hotspots , supporting our assumption that ISIS identifies hotspots . However , the limited amount of experimental data limits our ability to elaborate on this comparison . We also compared the conservation and the structural features of both groups . As shown in Figure 3 , there were hardly any differences in conservation . However , the most striking differences were found between structural features ( Table 2 ) . The secondary structure state of 39% of the non-hotspot interface residues was a loop . In the predicted hotspots , on the other hand , 57% of the residues were in a loop state . In both categories , the rest of the residues were divided roughly equally between helices and strands . Again , there is a striking agreement between the properties of predicted hotspots and the properties of experimental hotspots , despite the fact that ISIS was trained on all interface residues . Predicted hotspots were also much more accessible to solvent than other interface residues . Several studies suggested that hotspots have certain structural characteristics that differentiate them from other residues [37 , 38] . The Baker lab has shown that given a 3-D structure of a protein complex , it is possible to predict the results of alanine scans specifically and accurately [39 , 40] . This indicates that alanine scans indeed capture some genuine physicochemical commonalities of interaction hotspots that could be identified by a general method that is applicable to all protein complexes . The in silico alanine scanning is based on analysis of the 3-D structure of the interface between two proteins . Thus , it requires a high-resolution structure of the protein complex , while ISIS needs only sequence of a single chain regardless of its binding partner . On the other hand , in silico alanine scanning produces numerical prediction of the ΔΔG , While ISIS produces a binary prediction ( hotspot/non-hotspot ) . We compared our predictions to those of the in silico alanine scanning by translating their numerical predictions to binary ones according to cutoffs defined above . Of 55 experimental mutations with ΔΔG > 2 . 5 , in silico alanine scanning identified 36 ( 66% ) residues as hotspots . At this coverage , ISIS reached accuracy of about 60% while the in silico alanine scanning reached accuracy of greater than 75% . Scaled to an accuracy of 80% , ISIS identified 18 of these mutations ( 33% ) . Thus , for similar levels of positive accuracy , the coverage of ISIS is roughly half that of the in silico alanine scanning . Obviously , when structures of the complex are available , the in silico alanine scan is a powerful tool for identifying hotspots . However , when only the sequence is available , ISIS can provide accurate predictions for a substantial fraction of the hotspots . Our results indicate that some hotspots can be predicted accurately not only without relaying the 3-D structure of the complex but even without the 3-D structure of the unbound proteins . Furthermore , our predictions did not require knowledge of the binding partner . Analyzing a single protein using ISIS typically requires a few minutes . Thus , ISIS may allow large-scale analysis of hotspots at a relatively small CPU cost .
We used the ASEdb database of experimental alanine scans [12] , which lists residues that were mutated to alanine and the effect ( in terms of ΔΔG ) this mutation had on the interaction between two proteins . We checked the correlation between the predictions and the residues that were shown experimentally to substantially affect the affinity of the proteins in a complex to each other . In order to reduce the number of cases in which the effect of the mutation on binding was not due to a change in the interface ( e . g . , the cases in which the mutation destabilized the structure ) , we considered only exposed residues in proteins of known structure . Thus our test set included 80 protein chains with hundreds of experimental substitutions . From among these , we analyzed the mutations that substantially changed the binding energy ( ΔΔG > 2 . 5 kcal/mol ) , and those that had no effect ( ΔΔG = 0 ) . Altogether , we attempted to predict the experimental effect of 296 substitutions . The predictions were performed using ISIS [28] . ISIS can take as input either sequence or the coordinate of 3-D structure of unbound chains ( the results are more accurate when using known 3-D structures ) . However , for all values reported here , we ran ISIS from sequence alone . The accuracy and coverage of ISIS were measured using ratios derived from TP ( true positives ) , defined as the number of residues predicted by ISIS ( below ) to be in a protein–protein interface and observed to be in a hotspot ( i . e . , was found to have an extreme effect on binding; ΔΔG > 2 . 5kcal/mol ) ; FP ( false positives ) , defined as the number of residues predicted in protein–protein interfaces , were found however , upon mutation , to have no effect on binding ( ΔΔG = 0 ) ; and FN ( false negatives; i . e . , the number of residues predicted not to be in a protein–protein interface that were observed to have a strong effect on binding [ΔΔG > 2 . 5 kcal/mol] ) . We used the following equations: ISIS is a knowledge-based method we developed to identify interface residues from sequence [28] . It is based on a system of neural networks and uses as input the sequence environment of each residue , its evolutionary profile ( the frequency of each type of amino acid in a given position of the alignment ) , and its predicted secondary structure and accessibility to the solvent . In particular , when a sequence is submitted as a query , ISIS runs PSI-BLAST [41] , generates a multiple sequence alignment , and produces an evolutionary profile for each residue . These data are then sent to PROF [42] , a system of neural networks that predicts the secondary structure state and the solvent accessibility of each residue . Finally , the sequence environment , the evolutionary profile , and the predicted structural features serve as input to another neural network , which annotates each residue as interface or noninterface . ISIS was trained on a nonredundant version of all transient protein–protein interfaces [27] in the PDB . ( The 3-D structures were used only to identify the residues spatially in the interface . No experimental 3-D information was used for training . ) We trained standard feed-forward neural networks with back-propagation and momentum terms on windows of nine consecutive residues . A window was defined as positive if the central residue had any atom that was within 6 Å of any atom in a different protein . This yielded a set with 59 , 559 positive samples . We trained on two-thirds of the data and tested it on the remaining one-third . Next , we filtered the raw network predictions . Our analysis of protein interfaces at the sequence level suggested that most interacting residues have other interacting residues in their sequence neighborhood . Therefore , we eliminated predictions with fewer than seven raw predictions within ten adjacent residues ( five on either side ) . To obtain the expected coverage and accuracy at random , we reshuffled the predictions in the following way: each protein was represented by two strings of the same length , one representing its sequence and the other representing the predictions ( “P” for an interacting residue , “–” for a noninteracting residue ) . Then , we split the prediction string in half and assigned the predictions of the first half of the sequence to the second and vice versa . This process accounted for any size effect that could be caused by the number of predictions and for any effect caused by the heterogeneous distribution of contacting residues along the sequence . Furthermore , it enabled us to find a specific expectation for each scaling of the prediction . We generated different random models for different values of the receiver operating characteristic ( ROC ) –like curve ( Figure 2C ) . Our background model captured how random our predictions were rather than how well we could predict interface residues at random . ISIS was developed on a dataset of 1 , 134 chains in 333 complexes that contained 59 , 559 residue contacts . In the assessment of ISIS , no sequence that was used for training had any significant similarity for any of the sequences that were used for testing . That is , no protein in the test set could have been modeled by any protein in the development sets by homology-based predictions [43 , 44] . We chose methods that represent the variety of approaches for predicting interaction sites . ProMate [26] is a structure-based method that extracts features from an unbound chain and uses them to predict the binding site . We also chose three sequence-based methods: a sequence-only method [28] , an evolutionary-based method ( ET [35 , 45] ) , and a biophysics-based method ( hydrophobic moment [22] ) . The first two were available as servers for public use . The hydrophobic moment was not publicly available; thus , we implemented it for the purpose of this analysis . We chose an operating point of coverage equal to 15% , which was the highest coverage reached by the hydrophobic moment tool . We used the dataset of interface residues that was used to test ISIS originally [28] . In this dataset there are more than 20 , 000 interface residues , 2 , 182 of which were classified by ISIS as positive . Attempting to zoom in on the differences between hotspots and other interface residues , we compared the features of these 2 , 182 residues with the features of the residues that were classified as negative . The results of the comparison for amino acids are presented in Table 1 , and are based on the evolutionary profile we used for prediction . For each interface residue , we used a multiple sequence alignment to check how often each residue is present in this position . We performed the same analysis for all the positions that were found experimentally , by alanine scanning , to be hotspots . Table 1 shows the average percentage occupancy of each amino acid in all positively predicted positions in all negatively predicted interface residues . | Interactions between proteins underlie all biological processes . Hence , to fully understand or to control biological processes we need to unravel the principles of protein interactions . The quest for these principles has focused predominantly on the entire interfaces between two interacting proteins . However , it has been shown that only few of the interface residues are essential for the recognition and binding to other proteins . The identification of these residues , commonly referred to as binding “hotspots , ” is a first step toward understanding the function of proteins and studying their interactions . Experimentally , hotspots could be identified by mutating single residues—an expensive and laborious procedure that is not applicable on a large scale . Here , we show that it is possible to identify protein interaction hotspots computationally on a large scale based on the amino acid sequence of a single protein , without requiring the knowledge of its interaction partner . Our results suggest that most protein complexes are stabilized by similar basic principles . The ability to accurately and efficiently identify hotspots from sequence enables the annotation and analysis of protein–protein interaction hotspots in an entire organism and thus may benefit function prediction and drug development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biotechnology",
"biochemistry",
"none",
"eukaryotes",
"biophysics"
] | 2007 | Protein–Protein Interaction Hotspots Carved into Sequences |
The modular architecture of protein-protein interaction ( PPI ) networks is evident in diverse species with a wide range of complexity . However , the molecular components that lead to the evolution of modularity in PPI networks have not been clearly identified . Here , we show that weak domain-linear motif interactions ( DLIs ) are more likely to connect different biological modules than strong domain-domain interactions ( DDIs ) . This molecular division of labor is essential for the evolution of modularity in the complex PPI networks of diverse eukaryotic species . In particular , DLIs may compensate for the reduction in module boundaries that originate from increased connections between different modules in complex PPI networks . In addition , we show that the identification of biological modules can be greatly improved by including molecular characteristics of protein interactions . Our findings suggest that transient interactions have played a unique role in shaping the architecture and modularity of biological networks over the course of evolution .
Biological modules have played an important role in the evolution of cellular systems . After all , it is a group of genes , rather than a single gene , that cooperatively carries out cellular functions and determines phenotypic consequences [1] , [2] . Modules facilitate functional innovations in cellular systems , as modular rearrangements provide an efficient way to invent new cellular functions with a limited set of genes [3] , [4] . Moreover , modular architecture confers evolutionary robustness and stability to a system , by insulating it from the perturbing effects of genetic variation [5] , [6] . However , molecular-level understanding of the mechanisms underlying modular change in complex biological systems is currently not well developed . Current approaches to identifying modules in protein-protein interaction ( PPI ) networks often fail to consider the molecular components of connections . Hence , they cannot explain the molecular characteristics underpinning the evolution of network modules . Instead , they often rely on network topology , describing the organization of protein interactions [7]–[9] . Algorithms build topological clusters from protein interactions and try to identify clusters that correspond to certain biological modules , such as functional groups , protein complexes , and subcellular localizations . However , these approaches usually treat all interactions as equal and ignore differences in the nature of the connections . Social network studies have shown that network architecture and evolution are closely related to interaction strength [10] , [11] . Specifically , strong interactions , or long-term and intense commitments between people , are most likely to exist within communities ( Figure 1a ) . By contrast , weak interactions , or transient and distant acquaintances between people , tend to connect individuals in different communities . This pattern has an evolutionary origin: two unfamiliar people are more likely to develop a social tie and build a community if both of them have strong interactions to a common person [10] . Interaction strengths also influence how global networks function , including the rate and direction of information propagation [11] . Given that biological and social networks often share similar design principles , we anticipated that interaction strength would also affect the evolution of the modular architecture of biological networks . The physical characteristics of protein interactions are largely determined by their interface structures , which in general are classified into two groups: domain-domain interactions ( DDIs ) and domain-linear motif interactions ( DLIs ) [12] . DDIs usually display 103–106 fold stronger affinities than DLIs . Domains are globular structures of long peptides with defined binding or catalytic activities , whereas linear motifs are short peptides composed of specific sequence patterns that bind to other domains . Due to structural differences in the interacting components , DDIs tend to be characterized by large , strong interfaces between two globular domains , whereas DLIs are typically composed of small , weak interfaces between short peptides . In addition , domains and linear motifs have evolved in distinct manners . Domains are often conserved over a wide evolutionary range , evolving in a divergent manner [13] , whereas linear motifs tend to emerge from few substitutions in short peptides [14] , [15] . Therefore , we hypothesized that DDIs and DLIs may have made different contributions to the evolution of the modular architecture of PPI networks ( Figure 1b ) . In this study , we investigated the role of DLIs and DDIs in biological modules and found that DLIs are more likely to connect proteins between different biological modules , whereas DDIs tend to connect proteins within the same biological modules , including functional groups , protein complexes , and subcellular localizations . Furthermore , evolutionary analysis of PPI networks revealed that an expansion of DLIs in complex organisms has contributed to an increase in modularity , which may compensate for the cost of network complexity during evolution . We also demonstrated that module identification could be improved by utilizing DLI/DDI information . Indeed , interaction strength represents a unique biological aspect of network modules , one not incorporated by topology information alone . Our study suggests that inclusion of the physical characteristics of protein interactions will improve our understanding of the architecture and evolution of PPI networks .
We classified human PPIs into DDIs and DLIs to investigate the relationship between interaction strength and the modular architecture of networks ( Figure 2a; see Materials and Methods ) . Briefly , we categorized PPIs as DDIs if two interacting proteins had one or more domain-domain interactions . Interacting domain pairs were either identified directly from 3D structures of protein complexes [12] , [16] or from databases of domain-domain pairs [17] . We categorized PPIs as DLIs if two interacting proteins had one or more interacting domain-linear motif pairs . Interacting domain-linear motif pairs were identified from the Eukaryotic Linear Motif ( ELM ) database , which catalogs sequence patterns of linear motifs using regular expression and their interacting domains [18] . This procedure resulted in an integrated human PPI network containing 39 , 707 DDIs and 25 , 093 DLIs ( Table S1 ) . We found that the quality of linear motifs increased during DLI classification steps . Because linear motifs have high rate of false positives [18] , we assessed the fraction of true positive motifs in each step of DLI classification . A positive set of 695 experimentally validated motifs were collected from the ELM database and compared with randomly selected ones ( see Materials and Methods ) . We found that the fraction of true positive motifs significantly increased during the classification steps , especially , at the steps exploiting PPI neighbors to detect motif-binding domains and further removing overlap with DDIs ( Figure 2b ) . In contrast , the fraction of random sets remained unchanged during the steps . We also assessed the conservation of motifs since it has been reported that motifs involved in PPIs are relatively conserved [19] . We found that motifs selected from the classification steps are more conserved ( Figure 2b ) . Briefly , conservation score was calculated based on the information entropy of each column in multiple sequence alignments of orthologs and standardized over flanking residues ( see Materials and Methods ) . We further compared assigned DDIs and DLIs to reference sets in which the interfaces of human PPIs were identified directly from 3D structures or the literature ( see Materials and Methods ) . We found that assigned DDIs and DLIs accorded well with the reference sets ( Figure 2c ) . Specifically , 83 . 6% of the assigned DDIs ( n = 816 ) matched the reference DDIs , whereas only 1 . 0% of the assigned DLIs ( n = 10 ) were included in the reference DDI set . By contrast , 52 . 6% of the assigned DLIs ( n = 92 ) matched the reference DLIs , whereas only 1 . 7% of the assigned DDIs ( n = 3 ) were included in the reference DLI set . This also validates our approach to a classification of PPIs into DDIs and DLIs . We found that DDIs and DLIs have distinct roles in organizing the modular architecture of the human PPI network . DDIs tend to link proteins within the same topological clusters , whereas DLIs are more likely to connect different topological clusters in the network ( Figure 2a ) . To quantify this observation , we investigated the edge clustering coefficients of DDIs and DLIs ( see Materials and Methods ) . The edge clustering coefficient measures the fraction of connections between neighbors of two proteins connected by a given interaction [20] . Thus , interactions with a high clustering coefficient tend to connect proteins within the same topological cluster . We discovered that DDIs have higher edge clustering coefficients than DLIs ( Figure 2d , colored arrows ) . The average clustering coefficient of DDIs was 0 . 16 and that of DLIs was 0 . 061 ( Kolmogorov-Smirnov test , p = 1 . 0×10−323 ) . We confirmed that the observed clustering coefficients of DDIs and DLIs could not occur by random chance comparing them to randomly assigned ones ( Figure 2d , grey bars ) . The randomly assigned DDIs and DLIs were constructed by shuffling domains and linear motifs across proteins , while keeping the network connections unchanged ( see Materials and Methods ) . Note that false classification of DDIs or DLIs would lead the clustering coefficient similar to that of random ones because the network topology was not changed . The high clustering coefficients of actual DDIs and the low clustering coefficients of actual DLIs were significantly different than those of randomly assigned ones ( p = 1 . 0×10−5 for DDIs; p = 1 . 5×10−3 for DLIs ) . This was further confirmed based on the conservation of motifs constituting DLIs . We changed DLI datasets by varying motif conservation scores and measured average clustering coefficients . We found that the average clustering coefficients of DLIs were lower than that of DDIs , regardless of their motif conservation scores ( Figure S1 ) . Interestingly , the average clustering coefficients even decreased as the conservation of motifs increased . These indicate that the observed clustering coefficient would not likely emerge from false classifications . Because of the degeneracy in regular expressions , certain motifs could stochastically occur in many proteins . Therefore , we removed DLIs with low information content and reanalyzed the dataset . We confirmed that clustering coefficients of DLIs were lower than that of DDIs when we removed motifs with higher probability to be found by chance . DLIs showed lower clustering coefficient compared to DDIs even after removed 89 motifs with probability over 10−5 ( Figure S2a ) . Moreover , we found that the probability and clustering coefficient of motifs did not show significant correlation ( Figure S2b; p = 0 . 15 , Pearson's correlation ) . This confirms that DLIs generally have lower clustering coefficient , which is not restricted to several prevalent motifs . We next compared the role of DLIs and DDIs in various biological modules . Because biological modules are groups of proteins with tight functional relationships [1] , we investigated functional groups identified based on Gene Ontology ( GO ) terms . Protein complexes and subcellular localizations were also investigated , since they represent protein groups with particular functions [21]–[23] . We found that DLIs were enriched in protein interactions connecting different functional groups , whereas DDIs were enriched in interactions connecting proteins within the same functional group ( Figure 3a , Table S2 ) . Functional groups were identified using molecular functions ( MFs ) and biological processes ( BPs ) based on GO terms , while controlling for module size and overlapping relationships ( see Materials and Methods ) . For example , DLIs mediated by SH2 domains of Src kinase family proteins ( FYN , YES , LCK ) connect ‘cell-cell adhesion’ and ‘leukocyte migration’ protein groups ( Figure 3b ) . The Src kinases transiently dissociate p120-catenin ( CTNND ) and cadherins ( CDHs ) via phosphorylation , which results in short-lived gaps between vascular epithelial cells [24] . This enables leukocytes to transmigrate from blood vessel to tissue , which suggests that DLIs contribute to transient interactions between different functional groups . By contrast , DDIs connect proteins within the ‘cell-cell adhesion’ group through their Arm and Cadherin_C domains . And the proteins within the ‘leukocyte migration’ group are connected by the DDIs of the Pkinase_Tyr and Ras domains . We also confirmed that the bias of DLIs towards between-module interactions was observed regardless of their motif conservation ( Table S3 ) . We found that DLIs were enriched in between-complex interactions , whereas DDIs were enriched in within-complex interactions ( Figure 3c , Table S2 ) . For example , DLIs mediated by the BRCT domains of the BRCA1 protein connected the ‘RNA polymerase II’ and ‘BRCA1-associated genome surveillance’ complexes ( Figure 3d ) . The BRCT domain is a phosphopeptide-binding domain that mediates signal transduction events in the DNA damage response pathway [25] . BRCA1 interacts with the phosphorylated and functionally processive form of the RNA polymerase II complex to respond to DNA damage [26] , suggesting that DLIs contribute to transient interactions between different protein complexes . By contrast , DDIs connect proteins within the ‘RNA polymerase II’ complex via the TFIIE_alpha and BSD domains . In addition , the proteins within the ‘BRCA1-associated genome surveillance’ complex are connected by DDIs between the MutS and Helicase_C domains . We found that DLIs were enriched in protein interactions across different subcellular localizations , whereas DDIs were enriched in protein interactions within subcellular localizations ( Figure 3e , Table S2 ) . For example , the signal transducer and activator of transcription 3 ( STAT3 ) protein interacts with its partners in the cytoplasm and nucleus via DLIs ( Figure 3f ) . Specifically , the STAT3 protein transiently binds to heat shock protein 90 ( HSP90 ) in the cytoplasm and translocates to the nucleus , where it releases HSP90 to interact with other transcription factors [27] . By contrast , DDIs connect proteins with the same subcellular localization . For example , the Hsp70 and Hsp90 domains participate in protein interactions in the cytoplasm , whereas the Creb binding and Bromo domains participate in those in the nucleus . This suggests that DLIs contribute to the transient interactions of proteins that translocate between different subcellular localizations . We also provide more examples for the enrichment of DLIs and DDIs in interactions between and within biological modules ( Figure S3 ) . We confirmed that DDIs are biased toward within-module interactions regardless of they are mediated by same or different domains . One might ask that the observed bias of DDIs toward within-module interactions emerged from similar functions of identical domains . To test this question , we divided DDIs into two groups , homo- or hetero-DDIs . Any DDIs mediated by one or more pairs of same domains were classified as homo-DDIs and the rest of them were classified as hetero-DDIs based on their Pfam ID . We found that both homo- and hetero-DDIs are biased toward within-module interactions for functional groups , protein complexes , and subcellular localizations ( Table S4 ) . This indicates that the observed bias is likely due to the differences between DDI and DLI . Next , we investigated how the evolution of DLIs and DDIs contributed to the modular architecture of PPI networks . Comparative genomic studies have revealed that the number of peptide-binding domains and linear motifs , the basic components of DLIs , expanded as the complexity of organism increased [28] . We found the number of DLIs increased sharply in metazoan species ( Figure 4a; Table S5 ) . PPI networks for 45 nonmetazoan and 53 metazoan species were constructed using orthologous protein interactions from the human PPI network ( see Materials and Methods ) . Although the number of both DDIs and DLIs increased in metazoan PPI networks , the increase in DLIs was greater than that in DDIs . The average proportion of DLIs was 24 . 6% in nonmetazoan species; it increased to 40 . 2% in metazoan species ( Figure 4b; t-test , p = 2 . 4×10−43 ) . As expected , we found that the increases of linear motifs and DLI domains are more significant than that of DDI domains ( Figure S4 ) . What was the impact of this increased proportion of DLIs upon metazoan PPI networks ? We measured the modularity of PPI networks in eukaryotic species and found that the expansion of DLIs contributed to the modular architecture of metazoan PPI networks . To quantify the modularity of PPI networks in different species , we first applied a widely accepted topological measure , MPPI . By measuring the enrichment of within-module interactions , this measure was designed to assess to what extent modules are separated from each other ( see Materials and Methods ) . We discovered that the MPPI decreased sharply in metazoan PPI networks relative to those of nonmetazoans ( Figure 5a , Figure S5 ) . This decreased MPPI was due to an increase in between-module interactions , which connect proteins in different modules and reduce module boundaries ( Figure 5b , Table S5 ) . For example , the fraction of between-module interactions for protein complexes was 45 . 3% in nonmetazoans and 65 . 3% in metazoans ( Figure S6; p = 2 . 0×10−24 ) . We again tested whether the decrease of MPPI is due to any evolutionary association from same domains and found that MPPI decreased for both homo- and hetero-DDIs ( Figure S7 , S8 ) . Connections between different modules , however , do not necessarily reduce the modularity of PPI networks , because transient interactions between different modules are critical to the proper function of modular architecture . Therefore , we formulated a new modularity measure , MDLI/DDI , which takes into account DLI/DDI information; it incorporates the idea that DLIs mediate interactions between different modules , whereas DDIs mediate interactions within the same modules ( see Materials and Methods ) . In contrast to the decrease observed in the MPPI , we discovered that the MDLI/DDI increased in metazoan PPI networks relative to nonmetazoan networks ( Figure 5c , Figure S5 , S7 , S8 ) . Indeed , we found that DLIs tend to connect proteins at module boundaries , improving module quality in complex PPI networks ( Figure 5d ) . For example , novel Src family kinase ( FYN , YES , LCK ) DLIs emerged in metazoan species , regulating the transient opening of the junction between vascular epithelial cells in leukocyte migration [24] . Because of abundant connections between the two modular groups , each module's boundary is unclear at first glance . However , DLIs mediate the between-module connections of leukocyte migration and cell-cell adhesion modules , helping them cluster independently ( Figure 5e ) . Because DLIs and DDIs have distinct roles in the modular architecture of PPI networks , we employed DLI/DDI information in a topology-dependent module detection algorithm to improve identification of biological modules . We anticipated that DDIs would cluster proteins into modules , since they connect proteins with the same biological functions , whereas DLIs would separate proteins into different modules , since they involve transient interactions between proteins with different biological functions ( Figure 6a ) . To test this idea , we compared conventional topological PPI modules and DLI/DDI-identified modules . We constructed conventional PPI modules by using a greedy module-optimization algorithm , which consecutively merged single nodes to determine the architecture with the highest modularity ( see Materials and Methods ) . To construct improved modules , we applied DLI/DDI information by adjusting interaction weights . We found that considering DLI/DDI information dramatically improved the identification of biological modules ( Figure 6b ) . The quality of DLI/DDI-identified modules was significantly better than that of conventional PPI modules; this was true of various biological modules , including functional groups , protein complexes , and subcellular localizations . To quantify module quality , we analyzed the similarity of functional annotations , membership in protein complexes , and localization of subcellular compartments ( see Materials and Methods ) . The quality of functional groups was analyzed in terms of both MF and BP terms . We found that DLI/DDI-identified modules showed better quality than conventional PPI modules for various module sizes ( Figure S9 ) . Next , we investigated how DLI/DDI information could improve the merge process , resulting in better-quality protein clusters . By weighting network connections differently , the process prioritized the merging of DDIs in early steps and delayed DLI merges until later steps . For example , we found that voltage-gated Na+/K+ channel proteins ( HCN1-4 ) were grouped into the same module ( Figure 6c ) . A DDI between HCN2 and HCN4 ensured the merging of the two proteins in an early step . Conversely , DLIs between HCN proteins and Fcε signaling proteins ( FYN , SRC , GRB2 ) delayed the merge events for these proteins , resulting in separate modules . By contrast , based on conventional PPI information alone , HCN2 clustered with the FYN , SRC , and GRB2 proteins , becoming a member of the same functional module . This indicates that DLI/DDI information can improve the functional annotation process by identifying biologically relevant modules not easily identified using network topology alone .
In this study , we show that interaction strength plays a crucial role in shaping biological modules . Specifically , weak and transient interactions between modules promote the formation of functionally competent modular architecture in PPI networks , while a growing number of proteins and interactions have increased network complexity . Interestingly , it has been reported previously that weak interactions are enriched in between-module connections and are important for the proper function of various complex networks . For example , in social networks , weak interactions across community boundaries serve as passages along which novel information can travel [10] . Similarly , in the human brain , weak interactions connecting functional modules maximize information transfer at minimal wiring cost [29] . Indeed , interactions mediated by linear motifs are enriched in signaling and post-translational regulation networks [30] , [31] . This suggests that transient interactions mediating connections between modules may be a common design principle in complex networks . Thus , we propose that incorporating interaction strength into the study of network architecture provides novel insight into the principles of organization in biological systems . Due to the unstable characteristics , transient interactions are more difficult to detect than stable interactions [31] . We tested whether our conclusion is robust to underestimated transient interactions . Because multiple reports likely indicate more stable PPIs [32] , we constructed a stable PPI ( SPPI ) network using the PPIs found from two or more source of publications . We found that the clustering coefficient of DLIs was significantly smaller than that of DDIs ( Figure S10; p = 3 . 7×10−53 , u-test ) . We also found that DDIs and DLIs in SPPI network are enriched in within- and between-module interactions , respectively ( Table S6 ) . Therefore , we expect that our conclusions remain unchanged against future expansion of PPI networks with more transient interactions . We showed that DLI/DDI information can improve the identification of biological modules ( Figure 6 ) . Here , we focused on finding modules based on a conservative way , in which modules likely comprise strong DDIs between proteins with similar functions . Therefore , DLIs had been weighed lower than DDIs using a conventional framework which was designed to separate topological clusters . However , one might have another motivation of finding dynamic modules composed of transient interactions . We expect that DLIs and DDIs would also be informative in such cases because transient PPIs involved in dynamic cellular functions are likely mediated by DLIs [30] , [31] . One immediate way of finding dynamic modules would be to weigh DLIs over DDIs to find modules comprising DLIs rather than DDIs . This idea could be systematically tested when there were more experimental evidences for dynamic modules available from the advancement of detection methods for transient interactions [33] , [34] . We found that complex PPI networks displayed highly modular architecture when transient interactions were taken into account . Without proper consideration of transient interactions , however , complex PPI networks appeared to have lower levels of modularity than simple ones ( Figure 5 ) . It has been suggested that modular architecture is crucial in highly complex biological systems , to alleviate the “cost of complexity” during evolution [35] . For example , modules confer robustness to biological systems by insulating against the spread of perturbations originating from genetic variation . Without such insulation , perturbations could alter various functions , which would be likely to result in undesirable changes . Insulation becomes more critical as the complexity of biological systems increases; complex networks contain more components that can be perturbed than do simple ones [36] . In general , yeast and mouse experiments have shown that the effect of a single mutation is restricted , affecting a few traits [5] , [6] . This implies that modular pleiotropic structure does exist in the genotype-phenotype relationship . Our results highlight the fact that transient interactions are key in shaping the modular architecture of complex PPI networks . We found that DLIs mediate between-module interactions and that their relative abundance has dramatically increased in metazoan species . Functional innovations in metazoan species have often emerged from the rewiring of conserved functional modules [3] , [37] , [38] . Therefore , DLIs may be a key component of the rewiring of different functional modules in PPI networks . Indeed , linear motifs have been identified as “evolutionary interaction switches , ” because subtle amino acid changes can cause the short sequences in linear motifs to appear and disappear [14] , [15] , [39]–[41] . Furthermore , structurally disordered regions , where linear motifs are often located , have a high capacity for evolutionary rewiring in PPI networks [42] and largely increased in complex organisms [43] . This “switch-like” characteristic of short sequence motifs has been regarded as a prominent evolutionary mechanism affecting developmental processes in metazoan species . For example , mutations in cis-regulatory elements can selectively alter the expression of specific functional modules and result in dramatic changes in morphological patterns [44] , [45] . Our results suggest that subtle changes in short coding region peptides have also contributed to the rewiring of functional modules and , consequently , to functional innovations in metazoan species .
To assign DDI and DLI status , we first collected human PPI data from the following databases: the Human Protein Reference Database ( HPRD ) , release 9 [46]; BioGRID , release 3 . 2 . 107 [47]; IntAct [48] , downloaded December 3 , 2013; the Molecular Interaction Database ( MINT ) [49] , released March 26 , 2013; the Database of Interacting Proteins ( DIP ) [50] , released October 29 , 2013; Reactome v46 [51]; MatrixDB [52] , released August 1 , 2012; and InnatedDB [53] , released July 11 , 2013 . The integrated human PPI network comprised 264 , 845 interactions between 15 , 857 proteins . We classified a PPI as a DDI if two partner proteins had one or more interacting domain-domain pairs . Data on human protein domains were obtained from the Protein Family Database ( Pfam ) , release 27 . 0 [13] . Interacting domain-domain pairs were either identified directly from 3D structures or predicted using various computational approaches [17] . We first obtained 9 , 616 structurally characterized interacting domain-domain pairs from the Database of Three-dimensional Interacting Domains ( 3did ) , downloaded October 31 , 2013 [12] and iPfam , release 1 . 0 [16] , regarding them as the gold standard set . Then , every predicted interaction between domain-domain pairs received a confidence score:where CS ( i , j ) is the confidence score for the pair domain i and domain j , k indicates the prediction method , W is a precalculated weight factor for a specific prediction method , and I is an indicator of the prediction result ( Ik ( i , j ) = 1 if the method k gives a positive prediction for the pair domain i and domain j; Ik ( i , j ) = 0 otherwise ) . The weight factor assigned each prediction method was equal to its precision:where TP is the number of true positives , or the number of domain-domain pairs predicted by a given method and found in the gold standard set , and FP is the number of false positives , or the number of domain-domain pairs predicted by a given method but missing from the gold standard set . Predicted interactions between domain-domain pairs were considered valid if their confidence scores were greater than a cutoff value ( CS0 ) . To select a reliable CS0 , we investigated the F1 score of prediction results , increasing CS0 from 0 to 1 . 20 in 0 . 01 increments ( Figure S11 ) . The F1 score is the harmonic mean of precision and recall:where PR and RC are the precision and recall , respectively , of predicted interactions between domain-domain pairs with a CS>CS0 . Precision and recall were calculated as follows:where TP is the number of domain-domain pairs with CS>CS0 that were present in the gold standard set; FP is the number of domain-domain pairs with CS>CS0 that were missing in the gold standard set; and FN is the number of domain-domain pairs with CS<CS0 that were present in the gold standard set . Using the CS0 with the greatest F1 ( CS0 = 0 . 13 , F1 = 0 . 128 ) , we obtained 6 , 911 interacting domain-domain pairs predicted using various computational approaches . In total , this procedure gave us 16 , 527 interacting domain-domain pairs from both 3D structures and predictions . To avoid any bias in biological modules , we excluded prediction methods that exploited functional similarity . We classified a PPI as a DLI if two partner proteins had one or more interacting domain-linear motif pairs . We identified linear motifs in human proteins using regular expressions that represent motifs [18] . In contrast to other approaches , regular expressions have the flexibility to account for short indels and to provide presence/absence matches for motif patterns , simplifying the search . This feature is pertinent to our method , because interactions at the protein level will filter out most over-determined motifs . Two context filters provided by ELM server were also applied to the search . A taxonomic range filter removed linear motifs not related to human sequences . A structure filter removed linear motifs that overlapped with predicted secondary structures in globular domains . Interacting domain-linear motif pairs were obtained from “ELM classes” [18] . Each ELM class represents a pair of motif patterns and domains that interact with each other . Among the six types of ELM classes , we used ligand binding sites ( LIG ) , docking motifs ( DOC ) , and degron motifs ( DEG ) to focus on protein binding rather than the cleavage , targeting , or modification of motifs . PPIs remained unclassified if they satisfied criteria for both DDIs and DLIs . In total , we assigned 39 , 707 DDIs and 25 , 093 DLIs to 9 , 585 proteins . ELM instances , experimentally validated motifs in ELM database , were downloaded June 12 , 2014 . Among them , we found 695 positive and 12 negative motifs presented in the network . Because the number of negative motifs were too small to assess quantitatively , we also generated 10 , 000 random sets comprising 695 motifs of random selection for each and compared them to the positive set . We assessed the conservation of a motif using relative local conservation score ( RLC ) for each comprised residue and took their average for the motif [54] . RLC was calculated as follows:where CSV means conservation of residues from information entropy , μi and σi are mean and standard deviation of CSV , respectively , of [i−10 , i+10] residues including residue i itself . We used Shannon's entropy of each column in aligned ortholog sequences as CSV:where i denotes each column , α is an amino acid presented in a column , and P ( α ) is the frequency of the amino acid α in a column . Orthologs were obtained from Inparanoid database and only 100% confidence orthologs were used [55] . Otholog sequenes were aligned by MUSCLE algorithm [56] . For Figure S1 and Table S3 , DLIs were ordered by the highest conservation of comprising motifs and divided into different groups . We collected reference sets of human DDIs and DLIs whose status could be directly ascertained from 3D structures and literatures . Although 3did , iPfam and ELM databases provided experimentally confirmed DDIs and DLIs , only part of them might be interactions found in human proteins . Therefore , we chose reference DDIs from 3did and iPfam , if two protein constructs in the experiment were derived from human sequences by tracking species information from Protein Data Bank [57] . Reference DLIs were collected from ELM interactions by filtering out species other than human . Overlaps between reference DDIs and DLIs were discarded . The procedure resulted in 976 reference DDIs and 175 reference DLIs . Edge clustering coefficient measures the ratio of observed cyclic structures over possible cyclic structures around two connected nodes . Specifically edge clustering coefficient , C , between two nodes , i and j , was measured as follows [20]:where is the number of observed cyclic structures and is the number of possible cyclic structures among the partners of node i and j; g is the order of cycles , i . e . the number of nodes included in each cyclic structure . Here , we set g = 4 . We generated 10 , 000 permutations of DDIs and DLIs to obtain empirical p-values for the clustering coefficients . We permuted domains and linear motifs preserving their number in each protein and reassigned DDIs and DLIs . By definition , biological modules in PPI networks are groups of proteins that have tight functional relationships [1] . To determine functional groups of proteins , we used GO annotations , which provide a wide range of descriptions for the cellular function of proteins [58] . However , GO terms do not directly facilitate a clear division among functional groups , as they are designed to create hierarchical relationships in which parent terms include their child terms . To employ GO terms in a way that clearly separated functional groups , we first gathered certain GO terms with a comparable number of annotated proteins . We removed GO terms that displayed high levels of overlap , excluding the smaller of two GO terms when the union of the pair contained more than 50% of associated proteins . The procedure described was performed on terms from two functional GO categories: MF and BP . For protein complexes , we used the Mammalian Protein Complexes ( CORUM ) database [59] . We employed only human complexes , to prevent any bias originating from the higher level of conservation observed in DDIs [39] . Since several protein complexes with little variation can emerge from a subtle difference in the conditions employed in detection experiments , we removed those with high levels of overlap . As for functional groups , we excluded the smaller of two complexes whose union shared more than 50% of associated proteins . This procedure resulted in 1 , 217 protein complexes comprised of 2 , 646 proteins . We used the consensus localization prediction ( ConLoc ) method [22] to analyze subcellular localization . The algorithm first uses Universal Protein Resource ( Uniprot ) annotations , if available [60] . Then , it gives multiple predictions for subcellular localizations of a given protein , including associated confidence levels . In the cases in which no Uniprot annotation was available , we used the best prediction as the localization; we included the second prediction as well , if it was assigned over 80% confidence . This procedure resulted in 9 subcellular localizations for 18 , 575 proteins . To investigate the role of DLIs and DDIs in biological modules , we classified PPIs as within-module or between-module interactions . PPIs were considered within-module interactions if the interacting proteins had identical module memberships . Conversely , PPIs were considered between-module interactions if the interacting proteins had no common module membership . However , there were PPIs that met neither of these criteria ( dubbed “overlapping interactions” in Figure S12 ) . These overlapping interactions connected proteins that shared only part of their module memberships; thus , they could be interpreted either as within-module or between-module interactions . To be robust , we built two datasets . One treated overlapping interactions as within-module interactions , and the other classified overlapping interactions as between-module interactions . In both sets , our results were qualitatively similar , demonstrating that DLIs were enriched in between-module interactions and DDIs were enriched in within-module interactions ( Table S2 ) . Next , we further characterized the association of DLIs and DDIs with between and within-module interactions . We constructed a 2×2 contingency table with four types of interactions: between-module DLIs ( n11 ) , between-module DDIs ( n12 ) , within-module DLIs ( n21 ) , and within-module DDIs ( n22 ) . Enrichment was calculated as the observed number of interactions over the expected number of interactions for a specific association . For the observed number of nxy , the expected number was calculated as . For example , the expected number of between-module DLIs was ( n11+n12 ) × ( n11+n21 ) / ( n11+n12+n21+n22 ) , i . e . , the number of between-module interactions multiplied by the fraction of DLIs among the annotated proteins . We also determined if the level of enrichment was significant by calculating the p-value from Fisher's exact test . An analysis of MF terms for modules sized 80–160 proteins is shown in Figure 3 . We used protein orthology between human and other species to construct PPI networks and their modular architecture , as most interactomes were unknown when the genomes were sequenced . A human PPI was regarded as conserved in other species if the interacting pair of proteins had orthologs in them . Ortholog data were obtained from the Inparanoid database , and only 100% confidence orthologs were used [55] . Ortholog with the longest sequence was chosen , in case of multiple orthologs presented . To assign DDIs and DLIs , we searched domains and linear motifs in each species . To find domains , ortholog sequences were searched against the profile hidden Markov models of Pfam-A domains using pfam_scan . pl script and HMMER3 [13] , [61] . Linear motifs were searched using regular expressions and those overlapping with any domain region were discarded [18] . In this way , we constructed PPI networks for 45 nonmetazoan and 53 metazoan species . We used Newman modularity to measure MPPI [9] . The key assumption underlying topological modularity is that modules are separated from each other; the nodes within each module are densely connected , and the nodes between modules are sparsely connected . Specifically , topological modularity was calculated as follows:where lW is the number of interactions that connect proteins within the module , L is the number of interactions in the network , and dS is the sum of node degrees in the module . It measures the extent to which the proportion of observed within-module interactions exceeds the proportion expected by chance . However , MPPI strictly focuses on the separation of modules in network architecture , failing to recognize that biological modules influence each other . Indeed , the best MPPI score occurs when biological modules have no connection , which is unnatural . Given that DLIs likely connect different biological modules to carry out cellular functions , we revised MPPI to reflect that DDIs contribute to within-module interactions and DLIs contribute to between-module interactions . The revised modularity value , MDLI/DDI , was calculated as follows:where lWD is the number of DDIs that connect proteins within the module , lD is the number of DDIs in the network , lBL is the number of DLIs that connect proteins in the module to proteins outside the module , and lL is the number of DLIs in the network . The proportion expected by chance was adjusted for the proportion of DDIs and DLIs in the network . An analysis of BP terms for modules sized 80–160 proteins is shown in Figure 5 . To identify conventional PPI modules , we used a greedy modularity optimization algorithm [62] . Initially , each node was treated as a single module . Then , the algorithm merged nodes consecutively , until the entire network became a single module . In each step , all possible merge events between interacting nodes were evaluated by calculating changes in topological modularity , and the merge event with the greatest ( or least decreased ) value was selected . Modules were finalized according to the merged group of nodes with the highest modularity . Modules that possessed only two proteins were excluded from the analysis . We identified DLI/DDI-informed modules based on a procedure similar to the one used to identify conventional PPI modules; however , it weighted DDIs and DLIs differently [63] . In general , PPIs were categorized in a binary manner ( 1 if they existed , 0 if they did not ) . When an interaction was assigned to be DDI , its contribution to merging process is greater than a conventional PPI . By contrast , an interaction was assigned to be DLI , its contribution to merging process works in the opposite way . Thus , we weighted DDIs at 100 and DLIs at 0 . 1 . We used community_fastgreedy ( ) function in python-igraph package to build both PPI modules and DLI/DDI-identified modules ( http://igraph . org/python/ ) . The resulting modules were provided in Table S7 . We assessed module quality by measuring how similar proteins within the same module were . The similarity of each protein pair was calculated as the Jaccard index of biological annotations:where i , j is the protein pair and X is the set of biological annotations . Module quality was calculated as the average similarity of protein pairs . Fold increase in module quality was measured by comparing module quality to the average similarity of all protein pairs in the network . The p-value comparing module quality between the DLI/DDI-identified modules and conventional PPI modules was calculated using the Kolmogorov-Smirnov test . We also investigated the effect size of employing DLI/DDI information upon module quality using Cohen's d , designated e in Figure S8 . An analysis of MF terms for modules sized 80–160 proteins is shown in Figure 6 . | Modular architecture is important for the evolution of cellular systems . Modular rearrangements facilitate functional innovations and modular insulations provide robustness to perturbations . However , molecular-level understanding of the mechanisms underlying modular network evolution is currently not well understood . Here we show that strong domain-domain interactions ( DDIs ) and weak domain-linear motif interactions ( DLIs ) made different contributions to the evolution of the modular architecture of PPI networks . Especially , DLIs mediate between-module interactions , and that their relative abundance has dramatically increased in metazoan species . Linear motifs have been identified as evolutionary interaction switches since subtle amino acid changes can cause the short sequences in linear motifs to appear and disappear . Our results suggest that subtle changes in linear motifs have contributed to the rewiring of functional modules and , consequently , to functional innovations in metazoan species . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"protein-protein",
"interactions",
"protein",
"interactions",
"proteins",
"protein",
"structure",
"protein",
"domains",
"biology",
"and",
"life",
"sciences",
"molecular",
"evolution",
"evolutionary",
"biology",
"molecular",
"biology",
"macromolecular",
"stru... | 2014 | Linear Motif-Mediated Interactions Have Contributed to the Evolution of Modularity in Complex Protein Interaction Networks |
The survival of a bacterial pathogen within a host depends upon its ability to outmaneuver the host immune response . Thus , mutant pathogens provide a useful tool for dissecting host-pathogen relationships , as the strategies the microbe has evolved to counteract immunity reveal a host's immune mechanisms . In this study , we examined the pathogen Francisella novicida and identified new bacterial virulence factors that interact with different parts of the Drosophila melanogaster innate immune system . We performed a genome-wide screen to identify F . novicida genes required for growth and survival within the fly and identified a set of 149 negatively selected mutants . Among these , we identified a class of genes including the transcription factor oxyR , and the DNA repair proteins uvrB , recB , and ruvC that help F . novicida resist oxidative stress . We determined that these bacterial genes are virulence factors that allow F . novicida to counteract the fly melanization immune response . We then performed a second in vivo screen to identify an additional subset of bacterial genes that interact specifically with the imd signaling pathway . Most of these mutants have decreased resistance to the antimicrobial peptide polymyxin B . Characterization of a mutation in the putative transglutaminase FTN_0869 produced a curious result that could not easily be explained using known Drosophila immune responses . By using an unbiased genetic screen , these studies provide a new view of the Drosophila immune response from the perspective of a pathogen . We show that two branches of the fly's immunity are important for fighting F . novicida infections in a model host: melanization and an imd-regulated immune response , and identify bacterial genes that specifically counteract these host responses . Our work suggests that there may be more to learn about the fly immune system , as not all of the phenotypes we observe can be readily explained by its interactions with known immune responses .
The outcome of any infection , whether it be clearance of the infecting pathogen , establishment of a persistent infection , or even death of the host is determined by contributions from both the host and the microbe [1] . To infect a susceptible host microbes express virulence factors , genes that allow the pathogen to invade , colonize , and survive within the host and cause essential pathology . In response , the host initiates an immune response that attempts to clear the pathogen and increase tolerance to the ensuing infection [2] . Consequently , in addition to genes that allow the bacteria to invade host cells and obtain nutrients from its host , a subset of the virulence factors expressed by the microbe must address the need of the bacteria to counteract the host immune response . Exploring this complex interplay between host and pathogen can help us to understand bacterial pathogenesis and define the contributions of the host immune system to bacterial virulence . One way to explore the host-pathogen relationship is to apply model systems that allow us to dissect the genetics of both sides of the equation simultaneously in vivo [3] . In this study , we examine the host-pathogen interactions of Francisella novicida with an insect host , Drosophila melanogaster , and identify aspects of fly immunity that are most important for fighting F . novicida infection as well as the bacterial virulence factors that interact with each of these specific immune responses . Drosophila is used as a model of innate immunity because its simplicity and the ease at which it can be used for both forward and reverse genetics has allowed for the identification and characterization of aspects of the innate immune response that are conserved across evolution [4]–[6] . The fly immune response has three effector arms: an inducible antimicrobial peptide ( AMP ) response , a reactive oxygen response mediated by the activation of the enzyme phenoloxidase and the deposition of the pigment melanin , and a cellular immune response in which foreign invaders are phagocytosed by Drosophila hemocytes [7] . The humoral AMP response has been studied extensively and shown to be regulated by two pattern recognition pathways , Toll and imd which have been well-characterized and described , but the regulatory mechanisms of the melanization and cellular immune responses have only recently become the focus of increased interest and have not yet been fully elucidated [8] . Previous studies with pathogenic bacteria in the fly have shown that virulence factors that function in the vertebrate hosts of these pathogens are often required for the microbe to survive in the fly [6] . Recently , this has been shown to be true for the live vaccine strain ( LVS ) of the virulent pathogen Francisella tularensis , a Gram-negative facultative intracellular bacterium that is the causative agent of tularemia [9] . F . tularensis can infect a wide range of hosts that includes humans , but is more commonly associated with rabbits and small rodents . Unlike many of the pathogens used in previous fly studies , F . tularensis also has a documented arthropod vector phase in its natural life-cycle [10] , [11] . While many arthropod-vectored pathogens can only be transmitted by a single specific species , F . tularensis is able to infect arthropods ranging from ticks to multiple species of mosquito to biting flies such as deerflies [12]–[14] . This makes the Drosophila model system particularly useful for studying both general F . tularensis host-pathogen interactions and insect-specific factors . To date , the fly has primarily been used to dissect the function of known bacterial virulence factors or to demonstrate conservation between fly and vertebrate defenses . Less has been done to use forward genetic approaches in the microbe to identify virulence factors de novo . As immunologists we tend concentrate on known signaling pathways that have proven simple to study , are of interest to those working in vertebrates because they are conserved , and those that fit our idea of what the fly's immune response should be . In other words , experiments are driven by the interests of the scientists and not the pathogens . We took a more ecology-based approach and determined , from scratch , what this fly pathogen needs to kill the fly . The advantage of the fly is that it is inexpensive , rapid to use and has extensive genetic tools . The fly could be a useful tool for the identification of new virulence factors rather than a system used to study known factors . To identify new virulence factors and examine their interactions with the fly immune system , we used the Francisella novicida strain U112 to infect flies and performed a genome-wide screen to identify factors required for growth and/or survival within the fly . Many of the genes that we identified are required for resistance to the Drosophila innate immune response , particularly to the oxidative stress produced by melanization . This is interesting in particular , because until recently , this pathway had been discarded as having no relevance to microbial infections in the fly [15]–[17] . Our work demonstrates that bacterial mutants can be used as probes of the host immune system to identify what aspects of innate immunity are most important in determining the outcome of an infection . To identify additional interactions between the host immune system and bacterial virulence factors , we performed a second genetic screen in which we compared the ability of F . novicida mutants to grow in wild type flies to flies with an immunity defect known to affect fly survival in F . tularensis infections . These flies lack a functional imd signaling pathway , and we anticipated that this would reveal bacterial mechanisms necessary to circumvent the imd-regulated immune response . The imd pathway is primarily described as inducing antimicrobial peptides . Although we identified bacterial genes required to resist antimicrobial peptide killing in vitro , we were particularly intrigued to find a subset of genes that when mutated did not appear to change F . novicida sensitivity to the antimicrobial peptide we tested yet showed an altered phenotype in imd mutants . This suggests the possibility that there are previously undescribed immune mechanisms that are regulated by the imd pathway .
We infected flies with F . novicida strain U112 , a wild type strain that causes virulent infections in its natural mouse and rabbit hosts but is not pathogenic to humans . Previous work using the Live Vaccine Strain ( LVS ) of F . tularensis demonstrated that F . tularensis grows to high bacterial levels within flies and causes a lethal infection [9] . Infections of the fly with the U112 strain were consistent with this result , although we found the U112 strain to be slightly more virulent in Drosophila than the LVS strain , killing the fly with a median time to death ( MTD ) of 5 days post-infection with 103 CFU ( Figure 1A and Figure S1 ) . As few as 5 CFU of F . novicida U112 were sufficient to kill the fly , and bacterial growth within the fly was exponential approaching 5×107 CFU per fly before they succumbed to the infection . Regardless of the initial dose , F . novicida infections quickly reached the same high levels of bacteria; colony counts in flies receiving a low dose caught up to the 108 fold higher dose within two days ( Figure 1B and Figure S2 ) . In mammalian infections , F . tularensis is a facultative intracellular parasite that primarily grows within macrophages [18] . However , due to the extremely high bacterial levels observed within the fly we speculated that a large proportion of the bacteria were growing extracellularly . To test this idea , we performed gentamycin chase assays , infecting flies with F . novicida and subsequently injecting them with the non-cell permeable antibiotic gentamycin at various timepoints post infection . This assay kills off extracellular bacteria while leaving bacteria growing within cells intact and allowing us to determine whether the bacteria are growing intracellularly or extracellularly By 24 hours post infection , approximately 1×104 CFU per fly were found to be localized intracellularly . However , the average total population of bacteria in infected flies was at least 1×105 CFU per fly , demonstrating that a significant population of bacteria was located extracellularly . Over the course of the infection the total bacteria population increased to 1×107 CFU/fly , while the intracellular population remained steady , indicating that the extracellular population was responsible for the exponential increase in bacteria seen during fly infections ( Figure 1C ) . This is roughly similar to what has been observed using the LVS strain , although the absolute numbers differ slightly possibly due to the differences in virulence between the two bacterial strains , differences in the host strains or environmental conditions [9] . Having demonstrated that the fly can support F . novicida growth we applied this model to the identification of bacterial genes that were important for establishing infection within the fly . Previous work has shown that many F . tularensis virulence factors that are required for growth in mammalian models are also essential in insect infections , including the Francisella Pathogenicity Island ( FPI ) genes iglB , iglC , and iglD and the transcription factor mglA [9] , [19] . To expand upon this work , we sought to identify additional bacterial virulence factors and provide an opportunity to discover new biology using a forward genetic approach . Using a transposon insertion library of F . novicida mutants we performed an in vivo screen for mutants with altered growth rates compared to wild-type bacteria using a Transposon Site Hybridization ( TraSH ) assay . Briefly , flies were infected with the pooled library and the infection was allowed to proceed for two days , at which point the bacterial populations in each fly were harvested . Genomic DNA was then purified from this population of bacteria and from the original input library that had not been subjected to the stresses found within the fly . RNA was amplified from the site of each transposon insertion and the two populations of RNAs were compared by microarray analysis . We identified mutants representing 149 F . novicida genes that were negatively selected with a false discovery rate ( FDR ) of 5% , indicating that these genes were essential for bacterial growth and survival within the fly ( Table S1 ) . 41 of these genes had previously been identified as negatively selected in a similar TraSH analysis performed with the same mutant library in an in vivo mouse model; this list includes the known virulence genes iglC , iglD , pdpA , and pdpB , and mglA [20] . In addition , 11 genes from our screen overlapped with data from a negative selection screen performed by Kraemer et al . using an inhalation model to observe respiratory infections in the mouse , and an additional 8 overlapped with a signature-tagged mutagenesis screen performed by Su et al . that also used an intranasal route of inoculation [21] , [22] . The overlap between our TraSH assay and additional Francisella genome-wide screens is shown in Table 1 and Figure 2 . Interestingly , no genes were identified in all four unique screens , although 7 genes were identified in our fly TraSH and at least two other screens . These genes were the hypothetical proteins FTN_1682 and FTN_1016 , the RNA methyltransferase yibK , the ABC transporter yjjK , the amino acid antiporter FTN_0848 , iglC and iglD . The degree of overlap between our fly screen and similar mouse screens both supports the hypothesis that our screen in Drosophila can be used to identify virulence factors that are conserved between insect and mammalian infections , and also presents the possibility of identifying virulence factors unique to the arthropod vector stage of the F . novicida life cycle . To confirm the results of the TraSH screen , we tested 65 of the negatively selected mutants individually by competition assay , focusing on mutants that had particularly large decreases in abundance and/or showed homology to bacterial genes that we predicted could play a role in immune evasion or modulation . Transposon insertion mutants of each gene containing a kanamycin resistance cassette were tested to determine their ability to grow in competition with wild-type F . novicida . Each mutant was mixed with wild-type bacteria and injected into wild-type Drosophila at a 1∶1 ratio . Infected flies were incubated for 48 hours , at which point the bacteria in each fly were plated and the ratio of mutant to wild-type bacterial was determined . A competitive index of 1 represents an equal ratio of mutant to wild-type bacteria , while competitive indexes of less than one indicate that the mutant is attenuated . Mutants that were determined to be statistically significantly less than 1 by one sample t-test in a minimum of three repetitions were considered attenuated and are listed in Figure 2 . Mutants that were confirmed as negatively selected included kdpA , kdpC , kdpD , and kdpE , components of 2-component regulatory system that responds to turgor pressure , a number of genes known to be regulated by the virulence factor mglA , members of the Major Facilitator Superfamily ( MFS ) thought to be involved in substrate transport and drug resistance , multiple genes know to be involved in DNA repair , and a number of hypothetical proteins . ( Figure 3A and data not shown ) 56 of the mutants tested showed attenuated phenotypes by competition assay , with competitive indexes ranging from 0 . 6-0 . 007 . The results of the competition assays indicated that the microarray data produced by the TraSH assays is useful for predicting negatively selected mutants but was somewhat non-quantitative; the degree of attenuation as measured by microarray analysis did not always correlate with the strength of the phenotype observed by competition assay . One set of negatively selected mutants stood out as particularly interesting because they indicated a bacterial requirement for resistance to oxidative stress within the fly . These mutants included the LysR family transcriptional regulator oxyR . The homologue of this gene in E . coli has been shown to sense hydrogen peroxide and induce the transcription of downstream genes that provide protection against oxidative stress [23] . We also identified a number of genes that are essential for repairing damage to DNA such as that caused by reactive oxygen , including uvrA , uvrB , recB , ssb , mutM , and ruvC [24] ( Figure 3B ) . This result is consistent with a screen for attenuated F . novicida U112 transposon mutants using an inhalation method of inoculation , which identified the DNA repair proteins recO and recA [21] . The fly's melanization immune response produces reactive oxygen as an effector and thus we hypothesized that these bacterial genes were involved in helping F . novicida resist melanization [25] , [26] . To test this idea , we first performed in vitro disc diffusion assays to determine the sensitivity of each mutant to hydrogen peroxide ( H2O2 ) and paraquat [27] . The oxyR mutants were extremely sensitive to both H2O2 and paraquat ( Figure 3C ( H2O2 ) and data not shown ( paraquat ) ) . In addition , all of the DNA damage repair mutants showed a significant degree of sensitivity to both reactive oxygen-producing agents ( Figure 3D ) . Taken together , these data suggest that we identified a class of F . novicida genes that are essential for wild-type growth and survival within the fly , genes which help the bacteria to resist oxidative stress . Interestingly , 4 DNA repair genes , uvrA , recB , recO , and uvB , were identified in one screen of Francisella mutants in mice , suggesting that reactive oxygen species are a threat to the bacteria in mammalian infections as well . To determine whether melanization is a critical factor limiting the growth of F . novicida , we performed competition assays using the oxyR mutant in CG3066 mutant flies . These flies do not induce melanization upon infection [17] . We found that the growth defect of oxyR with respect to wild type bacteria was rescued in non-melanizing CG3066 flies ( Figure 3E ) . This supports the idea that melanization is the reactive oxygen producing immune response for limiting the growth of F . novicida in the fly . We therefore took our collection of negatively selected mutants and tested them for sensitivity to reactive oxygen . We found 25 mutants with increased sensitivity to oxidative stress and 2 with decreased sensitivity ( Figure 2 ) . Thus we were able to assign functions to these genes based on their behavior in a fly pathogenesis screen . This group of mutants makes up 45 percent of the mutants with attenuated growth phenotypes in the fly , demonstrating that oxidative stress mediated immunity is an important aspect of the fly's defenses against this pathogen and that an important class of F . novicida virulence factors in fly infections are genes that help the bacteria to counteract the effects of reactive oxygen . Having demonstrated that the TraSH method was useful for identifying genes required for growth in the fly and that many of these mutants were involved in counteracting the oxidative stress response of the fly , we decided to expand our study to examine an additional immune pathway and look for similar interactions . We chose to focus on one of the most intensely studied aspects of the fly innate immune response , NF-κB signaling pathways . Drosophila has two well-characterized NF-κB pathways ( Toll and imd ) that are responsible for sensing the presence of microbes and inducing an immune response [7] . Previous work by others demonstrated that the imd pathway is an important component of the Drosophila innate immune response to the LVS strain of F . tularensis , while the Toll pathway is not [9] . To confirm this for F . novicida U112 strain , we infected flies with null mutations in Toll and imd pathway genes . Two separate alleles of imd , imd1 and the null allele imd10191 , as well as mutants lacking the NF-κB homologue Relish showed significantly increased sensitivity to infections with the U112 strain; in contrast , mutants in the Toll pathway components Dif1 and dMyD88C03881 showed no significant difference compared to wild type ( Figure 4A ( imd alleles only , relish data not shown ) and Figure S1 ) . Therefore , we focused on F . novicida genes required to resist the fly's imd mediated response . To identify such genes , we repeated our TraSH analysis , this time infecting imd mutant flies and compared the results to those found in wild-type flies . We identified 36 genes that appeared to be negatively selected in wild type flies and at least partially rescued in imd mutant flies ( Table S2 ) Subsequent confirmation of these results by competition assay using transposon insertion mutants revealed a subset of 7 mutants that showed reproducible large rescue phenotypes in imd flies ( Figure 4C ) . These genes were the orphan response regulator pmrA , the gene FTN_0889 which is a helix-turn-helix protein and putative transcriptional regulator , glpD which is an anaerobic glycerol-3-phosphate dehydrogenase , the nicotinate-nucleotide pyrophosphorylase nadC , a uridine phosphorylase udp , FTN_0649 , a FAD-dependent 4Fe-4S ferrodoxin , and FTN_0869 , a hypothetical protein that encodes a putative transglutaminase that is regulated by the virulence factor mglA [28] . Since the imd pathway has been well-characterized as being responsible for inducing antimicrobial peptide ( AMP ) mRNA levels in response to F . novicida and other bacterial pathogens , the simplest explanation for the rescue of these bacterial mutants in flies lacking an intact imd pathway is that they have increased sensitivity to AMPs . This idea is further supported by identification of pmrA in our rescue screen , as pmrA has previously been shown to be sensitive to the antimicrobial peptide polymyxin B in vitro [29] . Therefore , we wished to determine whether the other F . novicida genes identified in our rescue screen are also sensitive to AMPs . There are 7 families of AMPs in Drosophila , and more than 2 dozen individual AMPs can be expressed during an infection , producing a complex bacteriocidal cocktail . Among the characterized AMPS , four families have been implicated in killing Gram-negative microbes , attacin , cecropin , diptericin , and drosocin . The first three families contain cation rich peptides while drosocin is described as proline rich [7] . It is currently impossible to recreate in vitro , the array of AMPs brought to bear on an infecting microbe in vivo . We therefore tried testing individual AMPs for their effects on F . novicida mutants . Unfortunately , few of these Drosophila AMPs are available commercially . We tested a commercial preparation of cecropin and did not detect activity against F . novicida on plates ( data not shown ) . We turned to the cationic antimicrobial peptide polymyxin B , which has been used to model AMP sensitivity in F . tularensis in multiple studies [29] , [30] . Of the seven genes we confirmed to be rescued in imd mutant flies , we found that five of these genes , pmrA , FTN_0889 , glpD , udp , and FTN_0649 were indeed more sensitive to polymyxin B in vitro . Suprisingly , mutants in the genes FTN_0869 and nadC did not show any phenotypes in these assays , suggesting that the imd rescue phenotype of these mutants may not be due AMP sensitivity , or at least not to cationic AMP sensitivity ( Figure 4D ) . To determine how common this phenotype was , we expanded our analysis to include the entire set of confirmed attenuated mutants described in Figure 2 . We found that twelve of the fourteen F . novicida mutants that were rescued in imd mutant flies on arrays showed altered sensitivity to polymyxin B , whereas this was the case with just five of the thirty eight mutants not rescued in an imd mutant ( Figure 2 ) . These five mutants were likely exceptions as they also had defects in reactive oxygen sensitivity and in the absence of an imd mediated response would still be sensitive to a melanization response . In our entire set of attenuated mutants , only nadC and FTN_0869 mutants demonstrated the unique phenotype of rescue in an imd mutant fly without showing any increased sensitivity to AMPs , so we chose to focus on one of these genes , the putative transglutaminase FTN_0869 for further analysis . The mutation in the gene FTN_0869 was intriguing as it clearly grows better in imd mutants as compared to wild type flies yet the mutant does not demonstrate altered sensitivity to the antimicrobial peptide we tested . The fly produces dozens of AMPs at once and not all of them work by the same mechanism , therefore it is impossible and illogical to eliminate the possibility that a single untested AMP or combination of imd induced AMPs might be responsible for killing F . novicida . Regardless , the resistance of FTN_0869 mutants to an AMP raises the question that the imd pathway might be generating an immune response that was not AMP mediated . In addition , the fact that this gene is regulated by the virulence factor mglA which regulates the F . tularensis pathogenicity island and many other important virulence factors suggested that it could be particularly important to F . tularensis pathogenesis . To determine the extent of the attenuation of FTN_0869 mutants , we examined the growth and survival of these bacteria in individual infections . We observed that with a starting dose of 5×103 bacteria the FTN_0869 mutant took significantly longer to kill wild-type flies than did wild-type F . novicida ( Figure 5A ) . This phenotype was completely rescued in imd mutant flies , with both FTN_0869 mutants and wild-type bacteria killing the fly with a mean time to death of 7 days , consistent with the sensitivity phenotype observed for imd flies ( Figure 5B and Figure S1 ) . In wild-type Drosophila , the FTN_0869 mutant did not develop the high bacterial loads found in wild type flies; wild type F . novicida can reach titers of 5×107 CFU per fly within 4 days while the FTN_0869 mutant did not grow higher than 5×105 CFU/fly ( Figure 5C ) . Again , this phenotype was abrogated in imd mutant flies , in which both wild type bacteria and FTN_0869 mutants were able to grow to similar high titers . ( Figure 5D ) This suggested that an imd-regulated immune response was preventing the FTN_0869 mutants from growing as well as wild-type U112 bacteria in the fly . We infer that the decreased bacterial population was responsible for the decreased virulence observed in terms of fly survival . The attenuated phenotype of mglA mutants in mouse cells is due to the inability of these mutants to survive and replicate intracellularly [31] . Since FTN_0869 is regulated by mglA , we sought to determine whether the same was true for this mutant . We performed gentamycin chase assays on wild-type U112 , mglA mutants , and FTN_0869 mutants . As expected , the mglA mutants showed no bacterial growth within the fly but rather were partially cleared very quickly following injection into the fly , and were completely unable to establish an intracellular population ( Figure 6A ) . This suggests that the small intracellular population may be important , if not essential , for the establishment of a successful infection . In contrast , the FTN_0869 mutants had a robust albeit slightly reduced intracellular population as compared to wild type bacteria , but demonstrated a unique phenotype with little to no extracellular bacteria present in wild-type flies ( Figure 6B ) . By testing sensitivity to gentamycin in vitro , we were able to show that this was due to lack of extracellular bacteria rather than an increased sensitivity of the FTN_0869 mutant to gentamycin ( Figure S3 ) . Again , loss of the imd pathway in the host animal eliminated this effect; the extracellular population of FTN_0869 mutant bacteria grew to similar levels as wild-type bacteria in imd mutant flies ( Figure 6C ) . This result suggested that the extracellular population of bacteria was unable to persist in the extracellular space of infected flies due to an immune mechanism that is controlled by the imd signaling pathway . Therefore , we were interested in investigating this mutant further to determine what effector arm of the innate immune system was responsible for the clearance of extracellular bacteria . We hypothesized that this clearance could be due to either an increased activation of the imd pathway by the FTN_0869 mutants , AMP activity that we were unable to test in vitro , or a novel component of the imd-regulated immune response . To determine if the imd pathway is induced more intensely by the FTN_0869 mutant , to rule out the possibility that this gene is able to downregulate the imd immune response , we measured the induction of imd-regulated AMPS as a readout for imd pathway activation . We used quantitative real-time RT-PCR to monitor the levels of Diptericin , Drosocin , Drosomycin , Attacin , Cecropin and Metchnikowin at 1 , 2 , 5 and 24 hours post-infection . We found that only Metchnikowin , Cecropin , and Diptericin were strongly induced in response to F . novicida infections and that the transcript levels of each of these highly-induced AMPs peaked at 24 hours post-infection ( Figure 7A and data not shown ) . All of these AMPs were induced to similar levels during infections with either the wild-type or FTN-0869 mutant bacteria , with no statistically significant difference between induction by wild-type or mutant bacteria at any timepoint . This confirms that the imd pathway is indeed activated by F . novicida and that the gene FTN_0869 does not have an effect on the induction of the imd pathway . We wished to probe the role of AMPs in clearing F . novicida further . There are more than 30 antimicrobial peptides in the fly and purified Drosophila AMPs are not readily available and the AMPs are always expressed together during an immune response; as described above , this makes it difficult to directly test the role of AMP activity on F . novicida growth in the fly . We therefore tried an indirect approach to test their importance . Recent work in the beetle Tenebrio molitor demonstrated that the majority of bacteria injected into the insect is cleared in less than an hour post-infection , much faster than antimicrobial peptides can be upregulated , transcribed , and synthesized [32] . Using this larger insect model , Haine et al . were able to conclusively demonstrate that insect antimicrobial peptide activity is induced slowly , and thus is not responsible for the bulk of the bacterial clearance . The analysis of antimicrobial peptide induction in the fly relies on the analysis of mRNA transcript levels , which are less accurate kinetically than a direct measurement of antimicrobial activity but nevertheless suggest that a slow induction with transcript levels only rising hours after infection and peaking at 6–24 hours post-infection for various AMPs [33] . To determine if the kinetics of extracellular bacterial clearance coincide with AMP induction , we performed gentamycin chase assays at early timepoints post infection . As early as 1 hour post-infection much of the FTN_0869 mutant population had already been cleared from the fly; in contrast the extracellular population of wild-type bacteria did not substantially decrease ( Figure 7B ) . By two hours post-infection , the timepoint at which both wild-type and mutant bacteria had begun to enter cells , the wild-type bacteria now had both intracellular and extracellular populations , while in FTN_0869 mutant infections only the intracellular bacteria had survived clearance . By five hours post infection the extracellular population of U112 wild-type bacteria had begun to increase while the titer of FTN_0869 mutants did not and only the intracellular population remained . This supported the notion that imd-induced AMPs were not responsible for the clearance of extracellular FTN_0869 mutant bacteria , as the bulk of this clearance occurred within an hour post-infection before AMP activity would be upregulated . We next sought to determine if one of the other effector arms of the fly innate immune system could be causing this phenotype . We first examined the effects of reactive oxygen species on the FTN_0869 mutants . Unlike many of the genes we isolated from our TraSH screen , the FTN_0869 mutants did not show increased sensitivity to reactive oxygen species produced by H2O2 or paraquat in vitro as measured by disk diffusion assay ( Figure 2 ) . We next examined the effects of melanization in vivo by infecting CG3066 mutant flies with wild type and FTN_0869 mutants . Unlike the oxyR mutants , the FTN_0869 mutants were just as attenuated in CG3066 mutants as they are in the wild-type control ( Figure 7C ) suggesting that these mutants do not have a defect in resisting reactive oxygen stress and that melanization is not responsible for the FTN_0869 imd rescue phenotype . We concluded that the imd rescue phenotype of FTN_0869 mutants not likely due to cationic antimicrobial peptides or melanization and rather suggested a third category of F . novicida interactions with the fly immune system as shown in Figure 2 .
Our goal was to dissect the host-pathogen interactions between Francisella and D . melanogaster . To identify components of this complex system we used a combination of three genetic techniques that enabled us to determine the contributions of both host and microbe to the virulence of the infection . First , we identified bacterial virulence factors necessary to infect the fly using a library of F . novicida mutants . Second , we used fly immunity mutants to confirm which host immune pathways were essential for fighting F . novicida infections . Finally , we combined these two techniques to identify subsets of bacterial virulence factors that allow the bacteria to counter-respond to specific immune attacks and evade immune clearance . This paper identifies genes from both the pathogen and the host that are components of each of these aspects of the host-pathogen relationship . To identify bacterial virulence factors , we performed an in vivo screen that identified 149 bacteria genes that are important for growth and survival within the fly . 41 of the 149 genes had previously been identified in a similar screen performed with the same bacterial library in the mouse indicating that many bacterial virulence factors are conserved between host species [20]–[22] . Genes that overlap between the Drosophila and mouse screens include known virulence factors such as mglA , iglC , and iglD , a number of various transporters , and some of the DNA repair genes we identified as helping F . novicida to survive oxidative stress . The remaining genes are unique to our screen performed in the fly model . These genes could either represent F . novicida genes that play a role specific to arthropod vectors , demonstrate a stronger phenotype in insects than in mammals , or were not identified in previous screens for experimental reasons . We note that of the 26 F . novicida mutants identified as being sensitive to reactive oxygen , 7 ( 27% ) had been previously identified as being important for virulence in vertebrates . In contrast , of the 16 mutants we found to be polymixin sensitive , only 1 ( 7% ) was identified previously as being important for virulence in vertebrates . The numbers in this study are small enough that differences in representation could be due to chance and therefore future work with more pathogens will be required to confirm the trends seen here; that said , analysis of interactions with the fly's reactive oxygen based immune response seems to be useful predictor of genes that will be of interest to those studying vertebrates . In contrast , analysis of the AMP and imd sensitive mutants is not as robust a tool for identifying mutations that will are relevant in vertebrates . Secondary screens of these mutants revealed important patterns that shed light onto what particular stresses F . novicida encounters within the fly . 25 of the 56 mutants that we confirmed to have reduced competitive indexes compared to wild-type F . novicida were also hyper sensitive to oxidative stress in vitro . This indicates that preventing or repairing damage caused by reactive oxygen species is an important survival strategy for F . novicida in insect infections . Of particular interest among the genes that were sensitive to oxidative stress was the gene oxyR , which has homology to an E . coli transcriptional regulator that senses and responds to the presence of hydrogen peroxide by inducing the transcription of catalases and other genes that can counteract oxidative stress 23 . In addition , our screen identified multiple genes in DNA damage repair pathways that are also sensitive to oxidative stress [24] , [25] . We expect that these genes are required to repair damage caused by reactive oxygen species to DNA as has been suggested by Kraemer et al [21] . Of the three effector arms that have been characterized in the immune response occurring within the fly's body cavity , the major producer of oxidative stress is the melanization response [15] , [17] , [26] . Therefore , we speculated that the large number of negatively selected bacterial mutants with oxidative stress sensitivity phenotypes suggested that the melanization response plays a large role in the fly's immune response to F . novicida . To test this hypothesis , we performed competition assays with the oxyR mutants in fly mutants that lack a melanization response . As expected , the attenuation of these mutants was rescued in flies that do not melanize and therefore would be expected to not produce toxic oxygen species . This demonstrated that melanization is an essential component of the fly immune response against F . novicida and demonstrated that we could use our characterizations of bacterial genes to learn about the fly immune system and understand the host-pathogen relationship . We note that reactive oxygen is a well-established immune effector in the Drosophila gut . Perhaps most microbes encountering the fly will face this immune barrier before encountering internal immune defenses . Thus protection against reactive oxygen is doubly important for fly pathogens [34] . Because microbes must withstand the host immune system to mount a successful infection , we were able to exploit the inherent ability of bacteria to function as metaphorical immunologists to identify which aspects of fly immunity were important to F . novicida infections . We next sought to determine if this system could be used in the reverse direction by manipulating the fly immune system to identify which bacterial virulence factors were responsible for interacting with one specific aspect of innate immunity . We did this by performing a second round of our in vivo screen for bacterial mutants in an immunocompromised fly . We focused on the imd-regulated humoral immune response , which had previously been identified as important for fighting F . tularensis infections [9] . We confirmed that the imd pathway , but not the Toll pathway , was essential in combating F . novicida infections , and performed our TraSH assay in imd mutant flies . We identified a subset of bacterial virulence factors that were important for infections of wild-type flies but not imd flies; this imd-regulated immune response has been primarily characterized for its role in the induction of antimicrobial peptides and therefore we tested these mutants for their sensitivity to a cationic , membrane active antimicrobial peptide , polymyxin B [29] , [30] . As expected , twelve of the fourteen mutants were sensitive to polymyxin killing in vitro , providing another example of how resistance to host immune responses is an important component of bacterial virulence . We identified 2 bacterial mutants that were not sensitive to polymyxin in vitro despite being rescued in imd mutants flies . This phenotype was unexpected as the majority of the literature suggests imd signaling drives antimicrobial peptide production and this is its most important job . We propose three explanations for this phenotype . First , the rescue phenotype could be due to specific sensitivity to additional antimicrobial peptides that were not tested in vitro; the bacteria show no sensitivity to polymyxin but could be sensitive to one of the 30 or more AMPs synthesized by flies . Second , the rescue phenotype could be due to the bacterial gene being an inhibitor of the imd pathway; in this case the bacteria would have wild type sensitivity to AMPs but would encounter increased concentrations of them in the fly because the bacteria could not inhibit AMP production . Finally , the rescue phenotype of these bacterial mutants could be due to an aspect of imd-regulated immunity that has not been previously described . To differentiate between these possibilities , we chose one imd-rescue mutant , the putative transglutaminase FTN_0869 to characterize further in terms of its interactions with the fly immune system . We chose this gene because it had a strongly attenuated phenotype in wild-type flies that was significantly rescued in imd mutants and because it has previously been shown to be regulated by the virulence factor mglA , which is essential for F . novicida intracellular growth 28 . More recently , the homologue of this gene in the extremely virulent Type A F . tularensis ssp . tularensis strain Shu4 was identified in a transcriptional analysis of genes that are upregulated inside mouse bone marrow-derived macrophages ( BMMs ) [35] . Interestingly , FTN_0869 deletion mutants in the less virulent U112 strain are unable to replicate in BMMs , but mutants of the homologue of this gene , FTT0989 in the SCHU4 strain did not demonstrate any intracellular replication defect [28] , [35] . Further characterization of the FTN_0869 mutants showed that these mutants are attenuated for both lethality to the fly and bacterial growth in an imd-dependent manner . However , unlike its transcriptional regulator mglA , the FTN_0869 mutant is capable of intracellular growth within flies , but is incapable of surviving in the extracellular space . This phenotype is consistent with what is observed in mouse bone marrow-derived macrophages with the virulent Shu4 strain , but not with the phenotype of FTN_0869 deletion mutants in the U112 strain . The reason for this difference is unclear , but it is interesting to note that the ability of the putative transglutaminase deletion mutants to grow intracellularly correlates with its virulence in mammalian and insect hosts . We found that the phenotype of FTN_0869 deletion mutants in flies is imd-dependent , and used this phenotype to investigate the role of the imd pathway in clearing the extracellular bacteria . With this mutant , we were able to show that the imd rescue phenotype of this particular mutant was not due to modulation of the imd pathway because AMP genes downstream of imd are induced to similar amounts in infections with wild-type and FTN_0869 bacteria . By examining the kinetics of the clearance of extracellular bacteria , we were able to limit the possibility that other imd-induced antimicrobial peptides that we did not test in vitro were causing the attenuation of the FTN_0869 mutant . Using non-melanizing mutants , we were able to rule out melanization as the cause of this phenotype , leaving us with the possibility that imd could either be regulating the cellular immune response or an uncharacterized effector arm of fly immunity . Thus the FTN_0869 phenotype suggested a third category of host-pathogen interactions between F . novicida and the Drosophila innate immune system . Future work with this mutant and other imd-rescue mutants identified in our screen could provide further insight into the biology of the imd-regulated fly immune response . In summary , reciprocal studies of a pathogen , F . novicida and a host , D . melanogaster , allowed us to identify genes in the pathogen required to counteract , evade , or resist host immune responses and allow bacterial growth and survival . These studies identified two branches of host immunity that are important for fighting F . novicida infections , melanization and imd-regulated immune responses and helped us to understand how the bacteria resists these responses . By identifying the mechanism of one or two bacterial mutants based on their sequence or interaction with fly mutants we developed assays to identify the mechanism of mutants with unknown function . Our work with one of these mutants , FTN_0869 , taught us that there is likely more to learn about the fly immune system as there are classes of F . novicida mutants that cannot immediately be explained by their interactions with the melanization response or AMPs . Our screen allowed us to pose directed questions and focus our investigations on particular aspects of the host immune system and the microbial strategies to evade this immune response , helping us to identify and characterize components of the host-pathogen relationship .
All experiments were performed in wild-type Oregon Red ( OR ) flies unless otherwise noted . The imd mutant fly line imd10191 is a null allele with a 26-nucleotide deletion at amino acid 179 that results in a frameshift mutation and has been backcrossed onto an OR background . The Toll pathway alleles tested in this study are Dif1 which is a complete loss of function mutant and MyD88C03881 The CG3066KG02818 Sp7 mutant flies are PiggyBack insertion mutants on a w1118 background ( Bloomington stock number 13494 ) , and w1118 flies are used as the wild-type control for these experiments . All experiments were performed on 5–7 day old age-matched male flies that were maintained on dextrose medium at 25°C and 65% humidity in a 12∶12h light dark cycle . Francisella novicida strain U112 was used for all experiments described in this paper . Bacterial stocks were grown in Tryptic Soy Broth ( TSB ) supplemented with 0 . 2% L-cysteine and cultured overnight under aerobic conditions at 37°C . Cultures were grown to an OD600 of 1 . 5–2 and diluted in PBS to OD600 0 . 005–0 . 01 for fly infections . Flies were anaesthetized with CO2 and injected with 50nL of bacteria using a glass needle and a Picospritzer III injector system ( Parker Hannifin ) . Each fly was injected in the ventrolateral surface of the fly abdomen and placed into fresh vials with no more than 20 flies per vial to prevent crowding . Following infection , the flies were incubated at either 25°C or 29°C as noted . Each survival curve was performed using 3 replicates of 20 flies each for a total of 60 flies per condition and each experiment was performed a minimum of three times . The number of dead flies was monitored daily and Kaplan-Meier survival curves were generated using GraphPad Prism software , and statistical analysis was performed using log-rank analysis . Individual infected flies were homogenized in 100µL of PBS , serially diluted , and plated onto Mueller-Hinton ( MH ) agar plates supplemented with 0 . 025% ferric pyrophosphate ( Sigma ) , 0 . 1% glucose , 0 . 025% calf serum ( GIBCO ) , and 0 . 02% Iso-VitaleX ( Benton Dickinson ) . Plates were incubated overnight and colonies were counted to determine the number of bacterial colony forming units ( CFUs ) per fly . Statistical significance was determined using unpaired two-tailed t-tests . Gentamycin chase experiments were performed as described about except that 50nL of 1mg/mL of gentamycin was injected into each fly 3 hours prior to plating [36] . Three sets of 30 flies were injected with 50nL of the trash library . Each fly received approximately 2*105 CFUs of bacteria , representing approximately 2-fold coverage of the library . The infection was allowed to proceed for two days at 29°C , at which point each fly was homogenized and plated onto MH agar . Plates were incubated at 37°C overnight , and the bacteria were collected and pooled and DNA was collected by phenol-chloroform extraction . Each pool was divided in half and digested with either BfaI or RsaI ( NEB ) and re-pooled to be used as a template for in vitro transcription with a MegaScript T7 Kit ( Ambion ) . The RNA was then purified and used for reverse transcription using a SuperScript III First Strand Synthesis Kit ( Invitrogen ) and random hexamer primers . The resulting cDNA was labeled with amino-allyl dUTP using Klenow ( exo- ) enzyme ( NEB ) . The input pool was then labelled with Cy5 and the day 2 pools with Cy3 and hybridized to Francisella microarrays as has been previously described [28] . Data was normalized using the Stanford Microarray Database according to the median log2 Cy5/Cy3 and filtered using a Cy3 net median intensity of 150 and a regression correlation of >0 . 6 . The dataset was then analyzed using SAM software using a blocked 2-class analysis to identify differences between the input and wild-type or input and imd mutant samples with a false discovery rate of 5% [37] . Genes that were selected for further analysis were knocked out of F . novicida individually to create deletion mutants . Briefly , 500bp of sequence 5′ and 3′ to the gene of interest was amplified from genomic F . novicida DNA using Phusion DNA Poylmerase ( NEB ) , and fused onto either side of a kanamycin cassette using a sewing PCR reaction . 38 The resulting PCR products were then transformed into chemically competent F . novicida U112 as described [28] and the mutants were confirmed by PCR . To confirm the bacterial growth attenuation phenotypes , 50nL of a 1∶1 ratio of mutant and wild-type bacteria at an OD600 of 0 . 01 was injected into flies . The infection was allowed to proceed for 2 days at 29°C , following which the flies were homogenized and plated onto MH agar plates with and without 30 mg/mL of kanamycin . Since only the mutant bacteria is capable of growing in kanamycin media , we were able to determine the number of wild-type and mutant bacterial CFUs for each fly by subtracting the number of mutant bacterial CFUs from the total CFUs per fly . A competitive index ( CI ) was determined using the formula CI = ( mutant CFU day 2/wild-type CFU day 2 ) / ( mutant CFU input/wild-type CFU input ) . To determine the sensitivity of various F . novicida mutants to oxidative stress and antimicrobial peptides , disk diffusion assays were performed using protocols adapted from Mohapatra et al . and Bakshi et al . [27] , [29] . Briefly , 50µL of overnight cultures of bacteria were plated onto MH agar plates to create a lawn of bacteria . Plates were allowed to dry for 10 minutes , and then 6mm Whatman filter paper disks ( Fisher Scientific ) were placed onto each plate and inoculated with 10µL of 100mM freshly diluted hydrogen peroxide ( Sigma ) or 10µL of a 10 mg/mL stock of polymyxin B . Plates were incubated overnight and the diameter of the zone of inhibition was measured for each sample . Three zones were measured for each mutant and each experiment was repeated three times . The fold increase of antimicrobial peptide expression following infection by wild-type and FTN_0869 mutant F . novicida was determined by isolating RNA from infected flies 6 and 24 hours post-infection by trizol extraction and performing qRT-PCR analysis using an iScript One-Step RT-PCR kit with SYBR Green ( Bio-Rad ) and a Bio-Rad icycler . The following primer sets were used: cecropin 5′ 5″-tcttcgttttcgtcgctctc-3′ , cecropin 3′ 5′-cttgttgagcgattcccagt-3′ , drosomycin 5′ 5′-gacttgttcgccctcttcg-3′ , drosomycin 3′ 5′-cttgcacacacgacgacag-3′ , diptericin 5′ 5′-accgcagtacccactcaatc-3′ , diptericin 3′ 5′-cccaagtgctgtccatatcc-3′ , attacin 5′ 5′-caatggcagacacaatctgg-3′ , attacin 3′ 5′-attcctgggaagttgctgtg-3 , drosocin 5′ 5′-ttcaccatcgttttcctgct-3′ , drosocin 3′ 5′-agcttgagccaggtgatcct-3′ , metchinkowin 5′ 5′-tcttggagcgatttttctgg3′ , metchnikowin 3′ 5′aataaattggacccggtcttg-3′ , ribosomal protein 15a 5′-tggaccacgaggaggctagg , 3′-gttggttgcatcctcggtga . | To infect a host and survive attacks from the host immune system , bacteria express genes that allow them to counteract immune responses . By identifying these genes we can learn how hosts fight infections and how bacteria resist immune attacks . We identified Francisella novicida genes that interact with the fruit fly immune system by performing a genetic screen of bacterial mutants . We identified genes that when mutated cause the bacteria to grow poorly within the fly . Many of these genes were shown to help the bacteria survive oxidative stress , providing resistance to an immune response known as melanization . We then identified bacterial genes that interact with another branch of the immune system , the imd pathway , by performing a second screen in imd mutant flies . We identified bacterial mutants that cannot grow in wild-type flies but are rescued in imd mutants , indicating an interaction with this pathway . We followed up one example from this screen and found that mutants in the gene FTN_0869 grow normally inside cells , but cannot grow extracellularly . We found that this was due to being unable to resist previously unexplored aspects of the imd-regulated immune response that help fight off F . novicida infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/bacterial",
"infections",
"microbiology/innate",
"immunity"
] | 2010 | Reciprocal Analysis of Francisella novicida Infections of a Drosophila melanogaster Model Reveal Host-Pathogen Conflicts Mediated by Reactive Oxygen and imd-Regulated Innate Immune Response |
Environmental transmission of the zoonotic parasite Toxoplasma gondii , which is shed only by felids , poses risks to human and animal health in temperate and tropical ecosystems . Atypical T . gondii genotypes have been linked to severe disease in people and the threatened population of California sea otters . To investigate land-to-sea parasite transmission , we screened 373 carnivores ( feral domestic cats , mountain lions , bobcats , foxes , and coyotes ) for T . gondii infection and examined the distribution of genotypes in 85 infected animals sampled near the sea otter range . Nested PCR-RFLP analyses and direct DNA sequencing at six independent polymorphic genetic loci ( B1 , SAG1 , SAG3 , GRA6 , L358 , and Apico ) were used to characterize T . gondii strains in infected animals . Strains consistent with Type X , a novel genotype previously identified in over 70% of infected sea otters and four terrestrial wild carnivores along the California coast , were detected in all sampled species , including domestic cats . However , odds of Type X infection were 14 times higher ( 95% CI: 1 . 3–148 . 6 ) for wild felids than feral domestic cats . Type X infection was also linked to undeveloped lands ( OR = 22 , 95% CI: 2 . 3–250 . 7 ) . A spatial cluster of terrestrial Type II infection ( P = 0 . 04 ) was identified in developed lands bordering an area of increased risk for sea otter Type II infection . Two spatial clusters of animals infected with strains consistent with Type X ( P≤0 . 01 ) were detected in less developed landscapes . Differences in T . gondii genotype prevalence among domestic and wild felids , as well as the spatial distribution of genotypes , suggest co-existing domestic and wild T . gondii transmission cycles that likely overlap at the interface of developed and undeveloped lands . Anthropogenic development driving contact between these cycles may increase atypical T . gondii genotypes in domestic cats and facilitate transmission of potentially more pathogenic genotypes to humans , domestic animals , and wildlife .
As human populations expand and alter global habitats , increased contact among people , domestic animals , and wildlife can lead to the emergence or re-emergence of diseases that threaten public and animal health [1]–[3] . Originally described in terrestrial environments , the zoonotic protozoan parasite , Toxoplasma gondii , is emerging as an important pathogen in aquatic systems . Toxoplasma gondii has been linked to widespread marine mammal infection and severe water-borne disease outbreaks in humans around the world [4] . The impact of T . gondii has been particularly significant in the threatened southern sea otter ( Enhydra lutris nereis ) population along California's central coast , where it has caused illness and death of otters of prime reproductive age [5] , [6] . In addition to playing a vital role in maintaining near-shore kelp forest ecosystems , sea otters serve as important sentinels for identifying disease threats , like T . gondii , to other marine species and humans sharing the coastal environment [6] , [7] . Warm-blooded animals and humans typically become infected with T . gondii by ingesting oocysts in contaminated water or food , by eating an intermediate host with parasite cysts in its tissues , or via vertical ( e . g . transplacental or transmammary ) transmission [8] . Oocysts , the extremely hardy , free-living stage of T . gondii [9]–[11] , play a key role in terrestrial and aquatic environmental transmission [12] , [13] . Epidemiologic association of sea otter T . gondii infection with diet , as well as laboratory and field evidence that marine invertebrates can filter and concentrate oocysts or collect oocysts while feeding on kelp , support the high likelihood of otter infection through oocyst-contaminated prey [14]–[18] . Because humans consume many of the same marine invertebrates , such as raw oysters , coastal contamination with T . gondii poses a public health risk [19] . As domestic and wild felids are the only known hosts capable of shedding T . gondii oocysts , high levels of marine mammal infection suggest land-to-sea transmission . Oocysts shed in felid feces become infective in the environment and can accumulate and survive for over a year in soil , freshwater , and seawater under a range of ambient conditions [20]–[22] . Environmental persistence increases the potential for oocysts to be transported in freshwater runoff to aquatic systems where sea otters , people , and other animals can become exposed . A terrestrial source of T . gondii is also supported by research linking increased sea otter T . gondii infection to coastal freshwater runoff [23] . Oocysts shed in the feces of pet and feral domestic cats , as well as wild felids , can serve as the source of sea otter and human infection , but contributions of these different felids to environmental oocyst contamination are not clearly defined . Molecular epidemiology offers a unique approach to trace pathogens from their sources to hosts in diverse environments [24]–[26] . Previous work has linked T . gondii genetics to infection patterns at both host population and landscape levels [27]–[29] . There is only one species of Toxoplasma , but different genotypes have been described based on the alleles present at multiple loci in T . gondii isolates [28] , [30] , [31] . Early population genetic analyses on T . gondii , primarily on isolates collected from domestic animals and humans in North America and Europe , identified three predominant , archetypal clonal lineages; Types I , II , and III [32] . However , in Central and South America , diverse atypical genotypes were later found infecting wildlife , domestic animals , and people [31] , [33]–[35] . Archetypal genotypes appear to dominate in humans and domestic animals in North America [31] , while atypical genotypes are more prevalent in wildlife [27] , [36]–[38] . Over 70% of T . gondii isolates from California sea otters sampled between 1998 and 2004 were identified as a novel , atypical genotype , named Type X [6] , [39] . Type X ( also designated Type 12 , Haplogroup 12; HG12 , and ToxoDB Genotype #5 ) and other closely related atypical genotypes commonly infect diverse wildlife species in the continental United States [27] , [31] . In coastal California , T . gondii strains consistent with Type X were previously found in three wild felids and one fox , but only archetypal genotypes were detected in the five domestic cats tested [16] . Separate , co-existing domestic and wild ( sylvatic ) cycles of T . gondii transmission may explain the distribution of archetypal and atypical genotypes in different hosts [40] , [41] . In French Guiana , these cycles have been linked to human influences on the landscape , with archetypal genotypes associated with domestic animals in anthropized or developed areas and atypical genotypes predominant in people and wildlife from undeveloped tropical rainforest [29] , [42] . To evaluate terrestrial sources of California sea otter T . gondii infection , and by proxy , potential sources of human exposure , we focused on two central research questions . 1 ) Do both domestic and wild felids contribute to the environmental transmission of Type X oocysts ? 2 ) How does human land use influence terrestrial and land-to-sea transmission of archetypal and atypical T . gondii ? Central coastal California offers an ideal environment to investigate spatial and species-based patterns of T . gondii infection . Sympatric wild and domestic felid populations share the landscape bordering the sea otter range , a mosaic of developed urban , rural , and agricultural lands and “wild” areas with minimal or no development . We aimed to identify T . gondii genotypes infecting a large sample of animals from sympatric populations of domestic cats , wild felids , and wild canids as well as risk factors for Type X infection and spatial clusters of archetypal and atypical T . gondii infection in the coastal environment . Although they do not contribute T . gondii oocysts to the environment , we included wild canids as sentinels of the genotypes circulating in the coastal lands within their limited home ranges . We use the types of T . gondii , risk factors , and spatial clusters identified in our analyses to explore local terrestrial cycles of T . gondii transmission and implications for human and animal health in temperate and tropical landscapes .
Feral domestic cats ( Felis catus ) , bobcats ( Lynx rufus ) , mountain lions ( Puma concolor ) , red foxes ( Vulpes vulpes ) , grey foxes ( Urocyon cinereoargenteus ) , and coyotes ( Canis latrans ) were opportunistically collected within 100 kilometers of Moss Landing ( 36°48'15 . 28" N , 121°47'13 . 09" W ) and Morro Bay ( 35°22'16 . 83" N , 120°51'25 . 00" W ) , California , USA , from 2006 through 2009 . These locations border marine sites previously identified as areas of increased risk for sea otter infection with T . gondii [5] , [6] , [17] , [23] , [43] . No animals were euthanized for the purpose of this study , and all animals were sampled through collaborations with regional animal shelters and state and federal wildlife protection programs . Carnivores , which were humanely euthanized or found dead ( additional detail in Table S1 ) , were submitted to the California Department of Fish and Wildlife for postmortem examination . The majority ( 90% ) of the free-ranging , unowned feral domestic cats were euthanized by animal shelter population control programs . These cats were identified as feral by study area residents who trapped and submitted them to the shelters or through behavioral testing conducted by trained shelter staff . Additional feral domestic cats and red foxes were trapped and euthanized as part of a conservation program to remove invasive predators from endangered shorebird habitat . Grey foxes were largely killed by cars ( 42% ) or euthanized by wildlife rehabilitation centers ( 47% ) . The majority of bobcats ( 85% ) and coyotes ( 88% ) died due to trauma ( predominantly caused by vehicles ) . Most mountain lions ( 93% ) were killed by motor vehicles or shot by wildlife wardens due to predation on domestic animals or due to public safety concerns . Based on postmortem examination , cause of death was not attributed to T . gondii in any of the sampled carnivores . Demographic data ( species , age , sex , and reported location of death or capture prior to submission to shelter or rehabilitation facility ) were recorded during postmortem examination . Brain and tongue tissue samples collected aseptically during postmortem examination were stored at −80°C prior to molecular testing . Aliquots of freshly collected brain tissue were stored in antibiotic saline at 4°C for parasite culture . To culture T . gondii tachyzoites , brain homogenate was inoculated onto uninfected cell culture monolayers as described previously [44] . Cultures were monitored for 30 days , and discarded if no parasite growth was observed . For cultures with parasite growth , tachyzoite-infected cells were harvested , and cell pellets were stored at −80°C prior to molecular testing . We used established nested PCR analyses , restriction fragment length polymorphism ( RFLP ) genotyping , and direct DNA sequencing at six polymorphic genetic loci ( B1 , SAG1 , SAG3 , GRA6 , L358 , and Apico ) to characterize the genotype ( s ) of T . gondii in infected hosts [16] , [39] . Although RFLP genotyping approaches have been developed to assess alleles at 11 loci , these methods have primarily been applied to T . gondii DNA extracted from isolates rather than tissue samples [30] , [45] . Due to the low burden of T . gondii bradyzoite cysts in many naturally infected carnivores , genotyping DNA extracted directly from tissues necessarily required us to focus on only a few , highly sensitive genetic loci [16] . However , performing DNA sequence analysis in addition to RFLP genotyping in the present study enhanced our ability to assess variation at target loci . For example , GRA6 sequences differentiate Type II and Type X alleles , which share a common RFLP pattern at this locus . Additionally , choosing loci that were used to characterize genotypes in many of the tested sea otters [6] , [39] , allowed us to make more consistent terrestrial to marine comparisons . For each carnivore , brain and tongue samples or culture-derived infected cell pellets were screened for T . gondii infection using nested primers targeting the 35 copy B1 locus [46] under previously described thermocyler conditions [47] . As T . gondii cysts in tissues of naturally infected animals can be non-uniformly distributed [16] , three separate tissue aliquots ( one brain , one tongue , and one randomly selected brain or tongue ) were tested for each terrestrial carnivore before classifying the animal as negative . Samples that were PCR-positive for T . gondii on B1 screening were additionally analyzed at single copy loci with nested SAG1 , SAG3 , GRA6 , L358 , and Apico primers [45] , [48] using previously described thermocycler conditions [39] . DNA was extracted from 25 milligrams ( mg ) of cryopreserved tissue or tachyzoite-infected culture cell pellet using DNeasy extraction kits ( Qiagen , Valencia , California , USA ) . The manufacturer's instructions were followed with one modification: AE buffer diluted 1∶10 and heated to 95°C was used in the final elution step . For all target genes , external primer 50 µL PCR reactions included 37 . 7 µL of DNase- and RNase-free , distilled water; 5 µL of PCR buffer ( 10X buffer containing 15 millimolar ( mM ) MgCl2 ) ; 4 µL of 2 . 5 mM dNTP mixture; 0 . 5 µL each of 50 micromolar ( µM ) forward and reverse primers; 1 . 5 units Taq Polymerase; and 2 µL of the DNA template . One µL of external primer amplification product was used in the nested reaction . Positive controls ( culture-derived and characterized T . gondii DNA ) and negative controls ( reagents only ) were included in each round of PCR . Nested amplification products were separated electrophoretically on a 2% agarose gel stained with ethidium bromide , and viewed under UV light . Products consistent in size with T . gondii-positive controls were digested with gene-specific restriction enzymes [39] , [45] at 37°C for one hour , and banding patterns were visualized under UV light following gel electrophoresis . DNA from PCR products was prepared with QIAquick gel extraction kits ( Qiagen ) and ExoSAP-IT ( Affymetrix , Santa Clara , California , USA ) and sequenced at the Division of Biological Sciences DNA sequencing facility ( University of California , Davis , USA ) . Distinctive RFLP banding patterns and DNA sequences were used to determine the alleles present at target loci [30] , [39] . RFLP patterns were used to identify Type I , Type II or III , and Type X alleles at the B1 locus . For B1 samples with atypical RFLP banding patterns and all alleles at the single copy loci , nucleotide sequences of T . gondii DNA were aligned and compared to consensus archetypal and Type X sequences using a Clustal W alignment algorithm in MEGA v . 5 . 05 [49] . Characteristic single nucleotide polymorphisms at established nucleotide positions [39] , [46] were evaluated to determine the type of the allele present at each locus . Samples with sequences identical to established archetypal or Type X sequences for a particular locus were classified as having a Type I , II , III , or X allele . Type A , a genotype that has been reported in sea otters and terrestrial wildlife , differs from Type X by a single nucleotide mutation at the GRA6 locus [27] , [50] . In order to effectively compare results from this study to previous molecular research on T . gondii in California hosts , which did not distinguish between Type X and Type A [6] , [16] , [39] , we consider “Type X” infection to represent both genotypes . For sequences with novel nucleotide polymorphisms at a target locus , the number of mutations to achieve the sample sequence relative to archetypal and Type X consensus sequences was considered . Sequences containing one or two nucleotide mutations compared to related archetypal or Type X consensus sequences were classified as genetically drifted alleles from the closest parent sequence . When more than two mutations were present or sequences were equally distant from established archetypal and Type X sequences based on numbers of nucleotide substitutions , alleles were classified as unique . The inheritance pattern of the alleles across the six independent , polymorphic loci was used to classify the type of T . gondii amplified from each infected carnivore . Associations between T . gondii genotype and demographic , temporal , and environmental variables were evaluated using logistic regression . Risk factors for infection with strains of T . gondii consistent with Type X versus strains with archetypal alleles or atypical allele mixtures were assessed in carnivores with T . gondii DNA amplified at the B1 locus ( Model 1 , n = 85 ) , as well as a more conservative subset of animals with alleles characterized at two or more loci ( Model 2 , n = 59 ) . Risk factors evaluated included carnivore group ( feral domestic cat , wild felid , or wild canid ) , age class ( juvenile or adult ) , sex , and season ( wet or dry ) and year of sampling . As T . gondii DNA was only amplified from one coyote , foxes and coyotes were grouped as “wild canids” . Similarly , due to limited sample sizes , mountain lions and bobcats were combined as “wild felids” . These intermediate host ( wild canid ) and definitive host ( wild felid ) groupings are epidemiologically reasonable as the pooled animals likely share similar T . gondii exposure histories . Although bobcats and mountain lions commonly consume different wild prey species [51]–[53] , previous studies in central coastal California found similar seroprevalences of infection and genotypes of T . gondii among these species [16] , [54] . To classify the season in which each animal was sampled , we used local daily rainfall data from the closest California Irrigation Management Information System ( CIMIS ) gauge station [55] . Animals collected on dates between the first and last precipitation events recorded from October to May , the time period with the majority of annual rainfall in the coastal study area , were considered “wet season” samples . Predominant land use ( developed vs . undeveloped ) within 5 km of the carnivore sampling location was also evaluated as a risk factor for Type X infection . For each carnivore , the street address or a detailed description of the location where the animal was trapped or found dead was recorded . These sampling locations were manually geocoded using Google Earth version 5 . 2 ( Google Inc , Mountain View , California , USA ) and mapped on publicly available California land use layers [56] , [57] using ArcGIS v . 10 ( ESRI , Redlands , California , USA ) . We used categories in the land use layers to identify developed and undeveloped areas of the coastal landscape . Developed areas included urban and rural lands with residential , commercial , industrial , and agricultural ( row crops , hay pastures , orchards , and vineyards ) development . Undeveloped areas had little to no human development and included forests , woodlands , grasslands , shrublands , and wetlands . Predominant land use at each carnivore sampling location was determined by buffering the sampling point by 5 km and assessing the proportion of developed vs . undeveloped habitat within the buffer . Two bobcats and four feral domestic cats were excluded from land use analyses due to lack of spatial data . Risk factors associated with T . gondii genotype in univariable logistic regression models ( P ≤ 0 . 20 ) were included in multivariable models . Potential confounders and interaction variables were evaluated and controlled for in the models as necessary . After incorporating biologically and statistically significant variables ( P ≤ 0 . 05 ) , Akaike's Information Criterion ( AIC ) was used to select a parsimonious multivariable model for each dataset . Final model fit was evaluated through graphical residual diagnostics . Adjusted odds ratios and respective 95% confidence intervals were estimated to evaluate the magnitude of the association between each variable and T . gondii genotype . Although most carnivores were collected from unique locations , there were nine sites where two animals were sampled in the Model 1 data . Model 2 data included five sites with two animals sampled . To account for potential unmeasured correlation among animals sampled at the same site , mixed effects logistic regression models with sampling location as a cluster variable were evaluated . Tests of within location correlation conducted with 5000 bootstrap replicates for each mixed effects model were non-significant ( P > 0 . 05 ) , indicating that conventional logistic regression models were adequate . Statistical tests were performed in R [58] with mixed effects model parameters estimated using the glmmML package [59] . Geographical clustering of T . gondii strains consistent with Type X and Type II was evaluated among infected carnivores ( animals with sampling location data and alleles characterized for at least two loci ) using a Bernoulli model elliptical scanning window , with a medium non-compactness penalty in SatScan v . 9 . 0 [60] , [61] . The default maximum spatial cluster size of 50% of the population at risk was used , and overlapping clusters were not permitted . Wild and domestic carnivores are not randomly distributed in the coastal landscape , and Type X and archetypal genotypes have been more frequently reported in wildlife and domestic cats , respectively . Therefore , the analyses were adjusted for type of animal sampled in order to identify geographic clusters of strains with Type II or X alleles rather than spatial differences in carnivore sampling [62] . A significance level of α = 0 . 05 was used for all spatial and non-spatial statistical analyses .
From 2006 through 2009 , 373 terrestrial carnivores were sampled along the central California coast . Toxoplasma gondii DNA was amplified at the B1 locus from 85 ( 23% ) of the sampled animals; 49 ( 30% ) of 166 feral domestic cats , 10 ( 14% ) of 73 mountain lions , 11 ( 41% ) of 27 bobcats , 14 ( 17% ) of 81 foxes , and one ( 4% ) of 26 coyotes ( Figure 1 ) . Parasite DNA was amplified from culture-derived , tachyzoite-infected cell pellets ( culture ) as well as brain and tongue tissue samples ( Table 1 ) . Archetypal and atypical alleles of T . gondii were detected , with Type X alleles identified at the B1 locus in 32 ( 38% ) of 85 T . gondii-infected animals . At B1 , Type X alleles were more prevalent in sampled mountain lions and bobcats than in feral domestic cats ( Figure 2 ) . Strains consistent with Type X remained more common in wild felids when the dataset was restricted to animals with T . gondii alleles identified at two or more loci . In addition to B1 Type X and archetypal alleles , unique atypical alleles were identified in 10 animals based on nucleotide sequence polymorphisms at established positions ( Table S2 ) . Genotype characterization of T . gondii from the majority of animals with unique alleles at the B1 locus was limited by lack of amplification for single copy loci . A fox and a feral cat were infected with strains of T . gondii with an allele pattern consistent with Type II isolates across loci , but each possessed a unique , non-Type II single nucleotide polymorphism ( SNP ) at SAG3 ( Table S3 ) , suggesting low levels of genetic drift . Additionally , one feral cat ( FC 1; Table 1 ) was co-infected with two T . gondii strains , one with Type I alleles and one with Type II alleles , which were amplified from tongue and brain tissue , respectively . All animals with T . gondii DNA amplified at six loci were identified as infected with Type II ( n = 17 ) or Type X ( n = 3 ) genotypes ( Table 1 ) . Characterizing the allele present at the L358 locus allowed us to distinguish between Type II and a Type 12 strain closely related to Type X [27] . One feral cat ( FC 29 ) with alleles determined at five loci was also infected with a T . gondii strain consistent with Type X . Four animals ( two feral cats and two foxes ) infected with strains with a mixture of alleles not consistent with archetypal strains or Type X were classified as atypical . The remainder of the T . gondii infected carnivores had DNA amplified at B1 and at least one additional locus ( n = 36 ) or the B1 locus alone ( n = 25 ) . While T . gondii genotype cannot be fully characterized based on the alleles at one or two loci , we chose loci ( B1 , SAG1 , GRA6 ) with distinct Type X RFLP patterns or sequence differences to help differentiate between Type X and archetypal strains . It is possible that some of the Type X and Type II strains identified based on fewer loci are actually atypical genotypes , but the likelihood that they were misclassified as Type X instead of archetypal is low . For the remaining results and the discussion , “Type X” refers to T . gondii strains with alleles consistent with Type X and “Type II” refers to T . gondii strains with alleles consistent with Type II . Demographic and environmental risk factors were identified for infection with T . gondii strains consistent with Type X . When all carnivores with alleles characterized at the B1 locus were included in a multivariable logistic regression model ( Model 1 ) , carnivore type was significantly associated with Type X infection ( Table 2 ) . The odds of infection with Type X T . gondii were almost five times higher in wild felids ( bobcats and mountain lions ) than in feral domestic cats . The odds of infection with strains consistent with Type X increased to 13 . 7 times higher in wild felids than feral domestic cats when a more conservative subset of carnivores , those with alleles identified for at least two loci , was examined ( Model 2 , Table 2 ) . The odds of Type X T . gondii infection in wild canids did not differ significantly from domestic cats in either model . Adjusting for carnivore group in multivariate models 1 and 2 , land use was a significant risk factor for Type X infection . Carnivores living in predominantly undeveloped lands were six ( Model 1 ) to 22 ( Model 2 ) times more likely to be infected with T . gondii strains consistent with Type X . All foxes infected by T . gondii strains consistent with Type X ( n = 6 ) lived in predominantly undeveloped lands near sampled mountain lions and bobcats with Type X strains . The majority of foxes with archetypal T . gondii infection ( 3 of 5 ) lived in urban and agricultural areas near domestic cats infected with these genotypes . Although Type X infection was more likely in feral domestic cats that lived in or closer to undeveloped lands , Type X-infected feral cats were also identified in heavily developed urban and agricultural areas . Season and year of sampling and age were not associated with Type X infection . No significant interactions among variables were identified . After adjusting for carnivore group , a significant geographic cluster of animals infected with T . gondii strains consistent with Type II based on alleles at two to six loci ( P = 0 . 04 ) was identified near the eastern edge of Monterey Bay ( Figure 3 ) . Type II-infected animals within this cluster included feral cats and foxes . This cluster remained significant when we tested a more conservative dataset in which animals were identified as infected with Type II T . gondii based on alleles characterized at all six loci . The terrestrial Type II cluster borders a previously established geographic cluster of Type II T . gondii infection in sea otters in Monterey Bay [27] . Genotype classification in these otters was based on alleles at B1 , SAG1 , SAG3 , and GRA6 . Two geographic clusters of carnivores infected with strains consistent with Type X were detected at the northern ( P<0 . 01 ) and southern ( P = 0 . 01 ) ends of Monterey Bay . Type X-infected carnivores within the northern cluster included bobcats , mountain lions , foxes , and feral domestic cats . No Type X-positive feral cats were found in the southern cluster . Previously reported locations for sea otters infected with strains consistent with Type X [6] , [16] , [39] border the two terrestrial Type X clusters . The Type II terrestrial cluster predominantly contained developed urban , rural , and agricultural lands , while the Type X clusters included a higher proportion of undeveloped lands . The four animals with atypical strains of T . gondii were all sampled within 10 km of each other in an area of heavily fragmented developed and undeveloped lands bordering central Monterey Bay ( Figure 3 ) .
Patterns of archetypal and atypical T . gondii infection derived from an opportunistic sample of feral domestic cats and wild carnivores may not perfectly reflect the true distribution of these T . gondii types at the population level . However , we used similar sampling approaches across species to create more accurate comparisons of T . gondii strains among carnivore groups . While logistical limitations prevented collection of a random sample of domestic cats and wild felids and canids , the majority of carnivores sampled ( 87% ) were apparently healthy animals killed by vehicles or humanely euthanized in population control or depredation programs ( Table S1 ) . Eighty-five percent of the subset of carnivores with T . gondii DNA detected also died due to these causes . While some wild canids , 9% of the total sampled animals , were euthanized due to illness in rehabilitation centers , T . gondii was not identified as a cause of death for any animals in this study . Type X T . gondii has been associated with severe disease in California sea otters [39] , but there is not any published evidence of increased virulence of this genotype in the sampled terrestrial carnivore species that would make Type X-infected animals more likely to be found dead or killed by cars than animals with archetypal infection . A strength of our approach is that domestic and wild carnivores were collected from temporally and spatially overlapping populations with animals from each group sampled in both developed and undeveloped habitats . All species were sampled in wet and dry seasons over the three year study period , with no association found between season or year of sampling and genotype of T . gondii infection . Given the similarity of opportunistic sampling among domestic cats , wild felids , and wild canids , it is unlikely that the collection method differentially biased the detection of archetypal and atypical genotypes among groups . Therefore , our approach allows robust comparisons of the distribution of strains consistent with Type X and archetypal genotypes among species and in the coastal landscape . Sympatric domestic and wild felids can both shed archetypal and atypical genotype oocysts , but limited sampling of these hosts often makes their contributions to environmental parasite load difficult to define [63] , [64] . Initial identification of strains consistent with Type X T . gondii in wild felids living along the sea otter range supported a sylvatic terrestrial source for sea otter exposure [16] . However , evidence of strains consistent with Type X infecting feral domestic cats in our study highlights their potential as a source of this novel genotype for sea otters . Using carnivore tissue samples limited our ability to amplify T . gondii DNA at the number of loci necessary to fully characterize Type X genotypes . However , Type X infection in sea otters and terrestrial carnivores has been classified based on RFLP and sequencing data at one ( B1 ) to four ( B1 , SAG1 , SAG3 , GRA6 ) loci [16] , [39] , making our results comparable to previous regional studies . Detection of strains consistent with Type X in 31% of infected feral domestic cats that had T . gondii alleles characterized at two or more loci indicates that domestic felids can also contribute to the burden of Type X oocysts reaching the near-shore marine environment via contaminated freshwater runoff . The distribution of genotypes in North American domestic cat populations is not fully defined , but Type X or closely related genotypes of T . gondii have also been identified in seven domestic cats from the southeastern United States [27] . The detection of T . gondii strains consistent with Type X in domestic cats in this study , but not in previous research in coastal California [16] likely reflects the larger number of domestic cats tested ( n = 168 vs . n = 5 ) . However , an increasing prevalence of Type X infection in feral cat populations may also have contributed to detection of Type X in these hosts . Large populations of domestic cats in the coastal California environment [54] may offset their lower prevalence of T . gondii Type X infection compared to wild felids , allowing these hosts to play a significant role in environmental accumulation of atypical as well as archetypal genotypes . Additionally , many domestic cats are concentrated in developed areas where conversion of natural habitats to impervious surfaces , like concrete and asphalt , facilitates pathogen transport in contaminated freshwater runoff to aquatic systems [65] . High levels of Type X T . gondii infection in wild felids and the predominance of archetypal genotypes in domestic cats living in the developed landscapes surrounding Monterey Bay strongly suggest that local domestic cycles driven by transmission of archetypal genotypes and independent Type X-based wild cycles co-exist in coastal central California . The proportions of archetypal and atypical genotypes , including Type X , detected in wild felids , wild canids , and domestic cats differed significantly . Mountain lions and bobcats were five to 14 times more likely to be infected with T . gondii strains consistent with Type X than feral domestic cats , which were predominantly infected with strains consistent with Type II . Reports of Type X in wildlife species from diverse areas of North America support a sylvatic origin and maintenance of this genotype [27] . The presence of domestic and wild T . gondii transmission cycles in coastal California is also supported by the differential distribution of genotypes across the coastal landscape . The majority of wild felids sampled in both developed ( urban , rural , and agricultural ) and undeveloped environments were infected with T . gondii strains consistent with Type X , whereas Type X infection in domestic cats was strongly linked to living in an undeveloped or “wild” area , where range overlap with mountain lions and bobcats was more likely . Toxoplasma gondii infection in foxes , which are intermediate hosts with small home ranges , offers additional insight on the genotypes present in the local environment . All foxes infected with strains consistent with Type X were sampled in undeveloped areas , whereas the majority of foxes with strains consistent with archetypal genotypes were collected from developed urban and agricultural areas near Type II-infected domestic cats . Statistically significant geographic clusters of animals infected with strains consistent with Type X were identified at the northern and southern ends of Monterey Bay , regions with predominantly undeveloped lands . In contrast , a cluster of Type II infection in the urban and agricultural lands bordering central Monterey Bay emphasized the dominance of archetypal genotypes in highly developed areas with large populations of pet and feral cats associated with humans . Emerging , severe atypical T . gondii infection in humans linked to wild felids in tropical jungles and archetypal strains circulating in sampled domestic animals provided evidence for the existence of similar wild and domestic transmission cycles associated with land use in French Guiana [29] , [41] . The existence of domestic and wild terrestrial T . gondii transmission cycles in different geographic areas of the California coast has the potential to strongly impact land-to-sea transmission and sea otter infection . The Type II T . gondii terrestrial cluster borders the marine area where almost all of the sea otters identified as infected with Type II were sampled [39] , suggesting that archetypal T . gondii oocysts shed by domestic cats and carried in freshwater runoff from these developed watersheds infected otters in the nearby ocean waters . Otters along the remainder of their range were predominantly infected with Type X [39] , [50] , and the presence of larger areas of undeveloped habitat suggested a link to Type X-infected wild felids . However , detection of T . gondii strains consistent with Type X in domestic cats indicates that they may also serve as a source of Type X oocysts infecting sea otters and other animals in the coastal environment , especially in areas where domestic cats share habitat with wild felids and are likely exposed to prey infected through a wild cycle of T . gondii transmission . The ability of pathogens to emerge in new hosts and environments can be linked to alterations of genes and ecosystems , and changes at both levels may impact human and animal health . Although many factors can contribute to the severity of toxoplasmosis in infected animals and people , T . gondii genotype may play a role in virulence [32] , [66] . While increased virulence of Type X in humans has not been established , other atypical genotypes have caused individual cases requiring intensive care and at least four recent and severe outbreaks of acute toxoplasmosis in humans in Canada , Brazil , Suriname , and French Guiana [40] , [41] , [67] , [68] . Three of these outbreaks were associated with eating wild game or ingesting oocysts from environmental sources , such as contaminated water , and two were linked to wild felids . The prevalence and genotypes of human T . gondii infection have not been investigated in our coastal California study area . Although the prevalence of human exposure to T . gondii in the United States decreased from the early 1990s to the early 2000s [69] , the waterborne outbreak of toxoplasmosis caused by an atypical genotype in coastal British Columbia , Canada in 1995 highlights the potential for atypical human infection to emerge in temperate environments [70] . Domestic cat infection with Type X and other atypical T . gondii strains therefore may increase animal and human environmental exposure to potentially more virulent strains in both temperate and tropical regions . Domestic and wild T . gondii transmission cycles that overlap at the interface of developed and undeveloped lands may increase local genotype diversity or introduce existing atypical genotypes to domestic hosts with potentially severe consequences for human and animal health [29] . Evidence of both of these processes was found using fully characterized isolates from wild and anthropized areas in French Guiana , but samples from these two environments were often separated by over 50 km . As the majority of our domestic cats and wild carnivores with T . gondii strains identified were sampled in close proximity at the interface of developed and undeveloped lands , our results offer a complementary detailed local view of potential interactions between domestic and wild cycles of T . gondii . The distribution of strains consistent with Type X and Type II in the developed and undeveloped lands bordering Monterey Bay ( Figure 3 ) provides evidence for movement of strains consistent with Type X into developed areas . Six domestic cats infected with strains consistent with Type X were sampled in highly developed urban habitats , which could indicate range and prey overlap with wild felids or emerging domestic cycles of Type X transmission . Although wild felids in California typically favor undeveloped environments [71] , [72] , mountain lions and bobcats sampled in urban and agricultural areas during this study and prior research [73] illustrate the potential for environmental overlap and indirect atypical T . gondii transmission by oocysts or infected prey to the domestic cats common in developed landscapes . Meiotic recombination following felid infection with two genotypes of T . gondii has the potential to generate new atypical genotypes [40] . For example , feral cat ( FC 49 ) was infected with an atypical T . gondii strain characterized by an admixture of Type X and Type II alleles . This domestic cat , which lived in a developed area close to undeveloped lands , was sampled within 5 km of a Type X-infected mountain lion . The unique mixture of alleles in this strain suggests that some domestic cats , and/or their prey base , are becoming co-infected with strains with Type X and archetypal T . gondii alleles . The four sampled carnivores ( two feral domestic cats and two foxes ) infected with unique atypical strains of T . gondii were collected with 10 km of each other in an area with a mixture of highly fragmented developed ( rural towns and agricultural areas ) and undeveloped lands . A mountain lion infected with a strain consistent with Type X and domestic cats infected with strains consistent with Type II were also sampled in this area , illustrating the potential for variation in diversity of genotypes even on a small spatial scale . Although the number of animals with atypical strains was small , their concentration in a fragmented mixture of developed and undeveloped lands could indicate that local landscape structure where domestic and wild T . gondii transmission cycles overlap can influence emergence of novel atypical genotypes . A mixture of fragmented developed and undeveloped lands in rural areas could increase overlap between domestic and wild cycles in several ways . Higher levels of T . gondii exposure are predicted in domestic cats living in rural areas due to changes in species composition of intermediate and definitive hosts as well as prey availability and domestic cat predation behavior along the urban-rural gradient [74] . Free-ranging domestic cats living in a fragmented rural landscape with increased agricultural-undeveloped edges may hunt more frequently and be more likely to contact wild felid oocysts or prey infected through the wild cycle than free-ranging domestic cats with access to cat food , human scraps , or garbage in urban areas with more discrete borders . Wild felids commonly use edge environments or riparian corridors in urban and agricultural areas , including California vineyard and orchards [75] , [76] . Fragmentation could also increase mountain lion access to and predation upon livestock , potentially exposing mountain lions to domestic cycle genotypes of T . gondii . These factors suggest that not only an interface between developed and undeveloped lands , but also the physical structure of the interface could drive overlap between domestic and wild T . gondii transmission cycles . Human development is rapidly reshaping temperate and tropical landscapes , and associated changes in pathogen transmission cycles could have significant impacts on public health and wildlife conservation [1] , [2] . In many tropical areas , rapid human population growth and agricultural expansion are driving conversion of undeveloped habitats into urban , rural , and agricultural uses . In the 1980s and 1990s , the majority of new agricultural lands for crops and livestock grazing in Central and South America and Africa were converted from intact and disturbed forests [77] . Human expansion into undeveloped areas can facilitate contact among domestic cats , wild felids , and sylvatic intermediate hosts of T . gondii , as well as human contact with oocysts from wild felids . Higher levels of overlap between these populations have the potential to increase human exposure to atypical and possibly more virulent strains of T . gondii [29] . In the case of California and the threatened sea otter population , land use change may expand the range of atypical T . gondii genotypes in human-dominated urban and rural landscapes and simultaneously increase the movement of oocysts to the ocean . Continued development of the terrestrial landscape in temperate and tropical areas may therefore increase the number of archetypal and atypical T . gondii oocysts and other pathogens flowing into freshwater systems and the near-shore marine environment , where they pose a risk to sea otters , other wildlife , and humans [6] , [23] , [78] , [79] . The molecular epidemiology approaches used to evaluate the linkages between T . gondii infection in domestic and wild hosts , as well as terrestrial and marine environments , have broader applications for a suite of other pathogens at the human-animal-environment interface . | Toxoplasma gondii , a global parasite shed by domestic and wild felids , can cause severe disease in people and animals . In coastal California , USA , many sea otters have died due to T . gondii . Because T . gondii is shed by felids on land , otter infection suggests that this extremely hardy parasite is transported in freshwater runoff to aquatic environments , where animals and humans can become exposed . Molecular characterization of T . gondii strains infecting terrestrial and marine hosts can provide clues about parasite transmission cycles and sources of otter infection . By testing 373 and characterizing T . gondii infection in 85 terrestrial carnivores ( domestic cats and wild carnivores ) sharing the California coast , we found that Type X , the type previously identified in over 70% of infected sea otters tested , was more common in wild felids than domestic cats . However , discovery of Type X in domestic cats in this region suggests that they may play an important role in marine infection , as their populations are larger than those of wild felids . Differences in types of T . gondii among carnivores and in urban and agricultural vs . undeveloped areas suggest that there are separate , but overlapping domestic and wild cycles of T . gondii transmission in coastal California . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases",
"veterinary",
"diseases",
"zoonoses",
"veterinary",
"epidemiology",
"veterinary",
"parasitology",
"medicine",
"and",
"health",
"sciences",
"epidemiology",
"parasitology",
"biology",
"and",
"life",
"sciences",
"spatial",
"epidemiology",
"protozoan",... | 2014 | Using Molecular Epidemiology to Track Toxoplasma gondii from Terrestrial Carnivores to Marine Hosts: Implications for Public Health and Conservation |
Elimination of onchocerciasis ( river blindness ) through mass administration of ivermectin in the six countries in Latin America where it is endemic is considered feasible due to the relatively small size and geographic isolation of endemic foci . We evaluated whether transmission of onchocerciasis has been interrupted in the endemic focus of Escuintla-Guatemala in Guatemala , based on World Health Organization criteria for the certification of elimination of onchocerciasis . We conducted evaluations of ocular morbidity and past exposure to Onchocerca volvulus in the human population , while potential vectors ( Simulium ochraceum ) were captured and tested for O . volvulus DNA; all of the evaluations were carried out in potentially endemic communities ( PEC; those with a history of actual or suspected transmission or those currently under semiannual mass treatment with ivermectin ) within the focus . The prevalence of microfilariae in the anterior segment of the eye in 329 individuals ( ≥7 years old , resident in the PEC for at least 5 years ) was 0% ( one-sided 95% confidence interval [CI] 0–0 . 9% ) . The prevalence of antibodies to a recombinant O . volvulus antigen ( Ov-16 ) in 6 , 432 school children ( aged 6 to 12 years old ) was 0% ( one-sided 95% IC 0–0 . 05% ) . Out of a total of 14 , 099 S . ochraceum tested for O . volvulus DNA , none was positive ( 95% CI 0–0 . 01% ) . The seasonal transmission potential was , therefore , 0 infective stage larvae per person per season . Based on these evaluations , transmission of onchocerciasis in the Escuintla-Guatemala focus has been successfully interrupted . Although this is the second onchocerciasis focus in Latin America to have demonstrated interruption of transmission , it is the first focus with a well-documented history of intense transmission to have eliminated O . volvulus .
Onchocerciasis ( river blindness ) is caused by a filarial nematode transmitted by black flies of the genus Simulium [1] . The disease may be mild ( dermatitis ) or severe ( visual impairment and blindness ) and is caused by the human immune response to microfilariae ( mf ) released by female adult worms as they move across subcutaneous tissue and spread throughout the body . Humans are the only known reservoir [2] . Onchocerciasis occurs throughout much of East and West Africa and Yemen , and was brought to the Americas through the slave trade [3] . It is now endemic to 6 countries in Latin America ( Brazil , Colombia , Ecuador , Guatemala , Mexico and Venezuela ) . Foci of transmission in the Americas are relatively small and geographically delimited compared to areas of transmission in Africa [4] . In part due to the geographical isolation of foci , the goal of the Onchocerciasis Elimination Program of the Americas ( OEPA ) is both to eliminate ocular morbidity throughout the region , and to permanently interrupt transmission where possible [1] , [5] . Control and eventual regional elimination of transmission is considered feasible due to the efficacy of ivermectin ( Mectizan® , donated by Merck&Co , Inc . ) as a microfilaricide , when used twice per year [6] , [7] . While ivermectin used in this manner prevents transmission of infections , it does not kill adult worms [8] although it may reduce their fecundity and lifespan [9] . OEPA , along with its ministry of health counterparts , supports mass treatment with ivermectin twice per year , with the goal of reaching 85% of eligible individuals ( those ≥5 years of age , ≥90 cm of height and ≥15 kg of weight; excluded are pregnant women and individuals with severe disease ) living in endemic areas . Recent reanalysis of information on the effectiveness of ivermectin delivered in this strategy has suggested that six and a half years ( 13 treatment rounds ) of such coverage can be sufficient to interrupt transmission [7] . Guatemala , with a population eligible for treatment of 175 , 881 ( to receive 351 , 762 treatments ) in 2006 [5] accounts for 38 . 5% of the endemic population eligible for treatment in Latin America . Guatemala has four endemic foci: Santa Rosa ( Department of Santa Rosa ) , Huehuetenango ( Department of Huehuetenango ) , Escuintla-Guatemala ( Departments of Escuintla and Guatemala ) and the Central Endemic Zone ( Departments of Suchitepéquez , Sololá and Chimaltenango; Figure 1 ) [10] . The Guatemalan Ministry of Public Health and Social Welfare ( MSPAS , in its Spanish acronym ) has been delivering ivermectin to endemic communities , through mass drug administration ( MDA ) , since 1988 [8] and has reached 85% of the eligible population at risk twice per year in all foci since 2001 ( Figure 2 ) [5] . Beginning in 2004 , the MSPAS , in partnership with OEPA , the US Centers for Disease Control and Prevention ( CDC ) and the Universidad del Valle de Guatemala ( UVG ) , began evaluating three of the four endemic foci in the country to determine whether transmission had been interrupted in these areas and if semiannual treatment could be suspended . Criteria for making these determinations are based on World Health Organization ( WHO ) guidelines presented in the 2001 document “Certification of Elimination of Human Onchocerciasis: Criteria and Procedures” [11] as adapted for field conditions by Lindblade et al . [12] and the OEPA steering committee ( the Program Coordinating Committee-PCC ) [1] . In summary , the criteria to be applied in areas with historically documented onchocerciasis transmission include: 1 ) demonstration of a prevalence of mf in the anterior segment ( MfAS ) of the eye ( anterior chamber and cornea ) to be less than 1%; 2 ) a cumulative incidence of O . volvulus infection of less than 0 . 1% in school-age children; and 3 ) a prevalence of infection in vectors of less than 0 . 05% [11] , [13] . Based on these criteria , Lindblade et al . demonstrated that transmission had been interrupted in the Santa Rosa focus; [12] subsequently , the PCC recommended to the Minister of Public Health of Guatemala that treatment be suspended in this focus . That recommendation was accepted and the Santa Rosa focus is currently under a three year post treatment surveillance phase to monitor for transmission recrudescence [5] . In this report , we evaluate the current status of transmission of O . volvulus in the Escuintla-Guatemala focus based on the adapted WHO criteria .
In 2007 , the Escuintla-Guatemala focus consisted of 49 , 616 individuals at risk , with 45 , 224 eligible for treatment , divided among 103 communities in the department of Escuintla ( 14 . 30°N , 90 . 79°W ) , and 14 communities in the department of Guatemala ( 14 . 62°N , 90 . 53°W ) . Historically , this focus included areas with intense transmission: between 1979–1982 , the community mf prevalence ranged from 8 to 38% . [14] A larval control effort from 1979–1989 significantly reduced both biting density and community mf prevalence . To ensure that all areas with current or past evidence of onchocerciasis transmission were included in this evaluation , all communities with at least one of the following characteristics were identified using historical data , including unpublished reports from the MSPAS and published articles: a ) past evidence of onchocerciasis transmission ( nodules or mf in at least one community resident ) ; b ) suspicion of past transmission suggested by a documented survey , which may not have found evidence of transmission; or c ) currently under semiannual ivermectin treatment by the MSPAS . A total of 155 communities satisfied at least one of these criteria ) ( Figure 3 ) . These potentially endemic communities ( PEC ) served as the sampling frame for all evaluations of the status of transmission of onchocerciasis ( Figure 3 ) . Because ocular morbidity is more likely to be found where onchocerciasis transmission is most intense , [15] we evaluated ocular lesions associated with onchocerciasis only in communities that historically had the highest rates of transmission . PEC with >0% nodule prevalence in the last 3 MSPAS surveys ( 1989 , 1990 and 1991 ) and elevation >800 m were considered in order to maximize the possibility of finding O . volvulus related morbidity . A total of 16 communities satisfied these criteria . Two of these communities were dropped before the evaluation began because they had less than 5 inhabitants , and one was not included because it effectively serves as a bedroom community for Guatemala City , making it very difficult to locate potential participants in their homes . The calculation of minimum sample size was based on estimating a population prevalence of MfAS of the eye of less than 1% . Finding 0 positive individuals out of 300 examined will allow a one-sided 95% confidence interval ( CI ) to exclude 1% . Given an estimated non-response rate of 10% , the total sample size required was 330 . All houses of the selected communities were mapped and the residents were censused using a pre-programmed hand-held personal digital assistant ( PDA ) with a global positioning system ( GPS ) attached . Eligible residents ( those who were ≥7 years old and who had resided in the community for at least the last 5 years ) were identified and recorded . As the ophthalmologist is capable of evaluating up to 90 individuals per day , we stratified communities into those with <90 eligible residents and those with ≥90 eligible residents . In the small communities ( <90 residents ) , all eligible individuals ( N∼266 ) were invited to participate . In the larger communities ( ≥90 residents ) , a PDA-based algorithm was applied in the field to randomly select 12% of the households and their members for inclusion in the evaluation ( N∼223 ) . An ophthalmologist ( OO ) with extensive experience conducting evaluations of onchocerciasis-related eye disease performed the ophthalmologic evaluations . Visual acuity was measured with a Snellen chart using standard methods . Ocular examinations were conducted with a split-lamp in a darkened area after the patients were asked to sit with their head between their legs for 5 minutes [16] . MfAS were noted as live/coiled or dead/straightened . Data were entered in the PDA and later downloaded to a database for subsequent analysis . We estimated the cumulative incidence of O . volvulus infection by measuring the prevalence of antibodies ( IgG4 ) to a recombinant antigen of O . volvulus , OV-16 , [17] in a stratified sample of school children 6–12 years old . We chose to stratify the sample into urban and rural schools because it was possible that levels of transmission would differ significantly between industrialized urban areas and rural communities located close to black fly breeding sites . The urban areas were taken to be the 2 large cities in the focus ( San Vicente Pacaya and Palín ) , and the remainder of the schools located outside these cities were considered to be rural . Information about schools and the number of children aged 6 through 12 who were attending was obtained with the help of the MSPAS and the Ministry of Education . Based on these figures , we estimated 4 , 674 eligible children in the urban schools and 9 , 815 in the rural schools . Schools were ordered at random within each stratum and then selected until the target sample size had been reached . The selected schools were visited and meetings were held with directors , teachers and parents to explain the evaluation . Teachers were asked to prepare a list of all enrolled children for the day of the evaluation . Based on the WHO certification for elimination criteria ( cumulative incidence <0 . 1% ) and considering antibody prevalence equivalent to the cumulative incidence rate , 3 , 000 children were required in each stratum to calculate a one-sided 95% CI that excluded 0 . 1% when no seropositives were encountered . Given an expected 30% non-response rate , our target sample size was 4 , 286 in each stratum . The methods used to collect finger-prick blood samples and data on residency from children participating in the evaluation have been described previously [12] . Briefly , each participant provided 80–120 uL of blood by standard sterile finger prick procedures . Whatman filter paper No . 2 was used to collect the blood directly after the finger prick . Children who didn't attend school on the appointed day were traced to their homes and asked to participate . Blood samples were processed within two months of collection using a standard ELISA [12] . Simulium flies were collected from November 2005 to April 2006 ( peak biting season ) in seven PEC and tested for infectivity in order to calculate the Seasonal Transmission Potential ( infected stage larvae per person per season , STP ) . Collection sites were selected through a rapid assessment of PEC to find those that satisfied the following criteria: a ) high densities of S . ochraceum; b ) presence of appropriate collection sites that capture areas where residents are most likely to be exposed to vectors ( i . e . casco , near a house and cafetal , near coffee plantations ) ; and c ) willingness of the owner ( in the case of fincas [plantations] ) or residents to participate . We used similar methods described by Lindblade et al . [12] . Two teams of two people ( collector and paid attractant , a male resident of the finca ≥18 years old ) rotated between two collection sites ( cafetal and casco ) in each PEC . The paid attractants were given ivermectin before starting collections and had finger-prick blood samples taken on filter paper at the beginning and at the end of the evaluation to evaluate exposure to O . volvulus . Collections started at 8:00 AM and ended at 5:00 PM taking 10 minute breaks at the end of every hour and a 1 hour break at noon . Each community was sampled two days per month . At the laboratory at the UVG , the heads and thoraces of S . ochraceum were separated from their bodies and up to 50 flies were pooled per tube , maintaining separate months and communities . The bodies were analyzed first using a standard polymerase chain reaction ( PCR ) assay to detect O . volvulus DNA [18] . Positives were confirmed by a second PCR . If a positive was confirmed , all the fly heads of that community were tested . The one-sided 95% confidence intervals for the prevalence of MfAS and antibodies to Ov16 were calculated using the SAS ( version 9 . 0 , SAS Institute , Cary NC ) FREQ procedure with the EXACT statement , BINOMIAL option and an alpha level of 0 . 10 . The Poolscreen 2 . 0 program was used to calculate the proportion of infective flies based on the number of positive pools [19] . Biting rates and STP were calculated according to standard methods [12] . All protocols received appropriate review and approval by the CDC ( Atlanta , GA ) , the ethics committee of the UVG ( Guatemala City , Guatemala ) , and the MSPAS ( Guatemala City , Guatemala ) . All participants ≥18 years of age or the parents or guardians of children <18 years old read or had read to them an informed consent form and then were asked to sign or mark with their finger to indicate their consent to participate . Children aged 7 to <18 years old were read or had read to them an assent form and asked to sign or mark with their fingerprint to indicate their willingness to participate . Paid attractants also read or had read to them a consent form and indicated their willingness to participate with their signature or fingerprint .
Of the 13 communities selected for the evaluation , one could not be reached due to road and weather conditions , and the only family in a second community could not be found on the day of the evaluation . We evaluated 329 ( 73 . 1% ) of the 450 eligible residents selected for inclusion in the evaluation . Of the total evaluated , 55% were women and 36% were 7–15 years old . Blindness due to onchocerciasis was not observed in any of the patients and 306 ( 93% ) individuals evaluated had their visual acuity measured in the range of 20/20–20/70 . No MfAS were found; the prevalence of MfAS was therefore 0 , with a one-sided 95% CI of 0–0 . 9% . In the urban area , we registered 4 , 674 enrolled children in 24 local schools , and 3 , 130 ( 67% ) participated in the evaluation . Due to an insufficient blood sample , the results for seven children could not be determined . Out of the 3 , 123 samples analyzed , there was 1 positive ( a 9 year old male living in an urban area ) for antibodies against O . volvulus . A second blood sample was requested and also tested positive . The sample was sent for additional testing at an experienced onchocerciasis laboratory in Mexico ( Instituto Politécnico Nacional , Reynosa , Mexico ) against recombinant antigens Ov10 , Ov11 and Ov16; the sample did not test positive for any of these antigens and we therefore concluded that it was a false positive and have recorded it as a negative result . In the rural area , we registered 4 , 614 enrolled children in 34 schools , and 3 , 316 ( 72% ) participated in the evaluation . A total of seven samples again had to be discarded due to insufficient sample . None of the 3 , 309 samples tested were positive for OV-16 . Therefore , the prevalence of antibodies to Ov16 in the Escuintla-Guatemala focus was 0 , and the one sided 95% IC was 0–0 . 05% . A total of 28 , 423 Simulium flies were caught in 1 , 320 hours of sampling , from November 2006 through April 2007 . None of the human attractants was found to be seropostitive for OV-16 either before or after the evaluation . Of the flies collected , 17 , 336 ( 61% ) were S . ochraceum and 11 , 087 ( 39% ) were S . metallicum ( not considered to be a vector of onchocerciasis when community mf prevalence is low [20] ) . The highest biting densities were measured in November and December . A total of 14 , 099 S . ochraceum in 303 pools were tested for O . volvulus DNA by PCR; all pools were negative , thus , prevalence was 0% and the 95% CI was 0–0 . 01% . The geometric mean biting rate for S . ochraceum was 11 . 0 bites/person/hour while the arithmetic mean daily biting rate flies was 177 bites/person/day . As the proportion of infective flies was 0 , the STP was also 0 . To calculate the maximum potential STP , we used the upper end of the 95% CI of the proportion infective and the geometric mean biting rate to calculate the maximum STP , assuming that each infective fly would have 1 infective-stage larva . The calculated maximum potential STP was 1 . 0 infective stage larvae transmitted per person per season .
Our findings , based on the ophthalmologic , entomologic and serologic evaluations adapted from the WHO guidelines for certification of elimination , indicate that transmission of O . volvulus has been successfully interrupted in the Escuintla-Guatemala focus . Our studies in this formerly endemic area demonstrated that the prevalence of mf in the anterior segment of the eye was less than 1% , evidence of active or prior infection ( or exposure ) as measured by antibodies to a recombinant O . volvulus antigen ( Ov-16 ) in school children was less than 0 . 1% , and O . volvulus DNA in vectors was under 0 . 05% ( with a STP of 0 infective stage larvae per person per season ) . O . volvulus transmission in the Escuintla-Guatemala focus was extensively documented from 1979 to 1984 by the Guatemala-Japan Cooperative Project on Onchocerciasis Research and Control , which conducted a large-scale larval elimination program in the area around the town of San Vicente Pacaya in the Department of Escuintla [21] . Several communities in that area had a prevalence of mf in the skin of 8% to 38% as recently as 1982 [14] . Nevertheless , community mf prevalence dropped from an average of 26% to 7% during the years of the larval elimination program [14] . Larval control efforts ceased in 1989 , and a MSPAS survey in1991 found community mf rates of 3% ( MSPAS , unpublished data; Figure 4 ) . No surveys had been conducted in Escuintla until the current report . However , the MSPAS provided semiannual ivermectin treatments in the focus , reaching more than 85% of the eligible population at risk twice per year from 2002 to 2007 ( Figure 2 ) . Santa Rosa was the first focus in the Americas to demonstrate interruption of onchocerciasis transmission , but there is evidence that transmission was extremely low to nonexistent prior to ivermectin distribution [5] . In contrast , levels of transmission in the Escuintla-Guatemala focus were historically higher than the Santa Rosa focus and transmission was well documented until at least the early 1980s . The larviciding efforts in the San Vicente Pacaya area from 1983–1989 were responsible for a significant decline in vector biting rates and , subsequently , community mf prevalence . The successful interruption of transmission after 12–13 rounds of MDA with ivermectin in San Vicente Pacaya may be at least partially due to the reduction in mf prevalence resulting from the larviciding campaign . While other areas of the Escuintla-Guatemala focus experienced a decline in mf prevalence without vector control , larviciding may be considered a complementary strategy to mass drug administration in areas of intense transmission . Although the reported specificity of the Ov-16 ELISA test is 90% , [17] our laboratory has now tested over 9 , 964 samples from endemic areas with only 1 false positive , a specificity of 99 . 99% . However , distinguishing false from true positives is challenging . We undertook extensive interviews of the family of the child initially found positive to rule out travel to other endemic areas or potential exposure to vectors during his daily activities . We tested other family members , including a grandfather who reported a nodule that was extirpated in the past , and none was found positive . After a second blood sample from the same child tested positive , we sent the samples for testing in another laboratory against additional antigens , where the child's samples were negative in all external tests . We , therefore , feel confident reporting this finding as a false positive . The data presented in this report were extensively reviewed by OEPA and the PCC . Based on the results , the PCC recommended to the Minister of Health of Guatemala that ivermectin treatments be suspended in the Escuintla-Guatemala focus in 2008 . The recommendation was accepted , and , as in Santa Rosa , three years of surveillance for recrudescence has now begun during which a final set of evaluations to ensure that transmission has been completely eliminated will be undertaken [5] . Currently we are conducting a similar series of evaluations in the focus of Huehuetenango along the border with Chiapas , Mexico , to determine whether transmission has been interrupted there . Results from these studies are expected mid-2008 . As of the writing of this article , transmission of O . volvulus continues in the Central Endemic zone of Guatemala [5] . | Brought to the Americas from Africa by the slave trade , onchocerciasis is present in six countries in Latin America . The disease is caused by a round worm and is transmitted to humans by the bite of an infected black fly . Once in a human , the adult worms produce larvae that circulate through the body , causing itching or even blindness . Ivermectin , a drug that kills the larvae , is delivered by public health authorities in countries where the disease is present . If the larvae are killed , then the disease cannot be transmitted to more people . People living in the Escuintla-Guatemala focus , a region in Guatemala where the disease was common , have been taking ivermectin for many years . The Ministry of Health of Guatemala believes that onchocerciasis is no longer being transmitted in the area . To prove that there is no more transmission of the disease , the authors examined the eyes of residents of the area to see if they could find any evidence of the worms . They also conducted analyses of blood in school children to see if they had ever been exposed to the worm , and they caught thousands of black flies and tested them to see if they were infected . These evaluations found no evidence of transmission of the disease in the Escuintla-Guatemala focus . As a result , local public health authorities can stop giving ivermectin and invest their human resources in other important diseases . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/helminth",
"infections",
"public",
"health",
"and",
"epidemiology/global",
"health",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases"
] | 2009 | Successful Interruption of Transmission of Onchocerca volvulus in the Escuintla-Guatemala Focus, Guatemala |
Snakebite envenoming is a major public health burden in tropical parts of the developing world . In sub-Saharan Africa , neglect has led to a scarcity of antivenoms threatening the lives and limbs of snakebite victims . Technological advances within antivenom are warranted , but should be evaluated not only on their possible therapeutic impact , but also on their cost-competitiveness . Recombinant antivenoms based on oligoclonal mixtures of human IgG antibodies produced by CHO cell cultivation may be the key to obtaining better snakebite envenoming therapies . Based on industry data , the cost of treatment for a snakebite envenoming with a recombinant antivenom is estimated to be in the range USD 60–250 for the Final Drug Product . One of the effective antivenoms ( SAIMR Snake Polyvalent Antivenom from the South African Vaccine Producers ) currently on the market has been reported to have a wholesale price of USD 640 per treatment for an average snakebite . Recombinant antivenoms may therefore in the future be a cost-competitive alternative to existing serum-based antivenoms .
The global disease burden from snakebite envenoming is massive , and particularly affecting poor rural tropical areas in Africa , Asia , Oceania , and Latin America [1] . The incidence of envenoming is estimated to be in the order of 2–3 million per year , resulting in more than 100 , 000 deaths [2 , 3] . Although animal-derived antisera remain the cornerstone of snakebite therapy [4] , biotechnological advances are driving the emergence of different antivenom formats based on human or camelid antibody scaffolds [5 , 6] , which in the future may pave the way for recombinant oligoclonal mixtures of antivenom antibodies [7] . The potential benefits of recombinant antivenoms for treatment of snakebite envenoming include higher potency and fewer side effects ( serum sickness and anaphylaxis is not uncommon from animal-derived antisera ) due to the possibility of producing fully human antibody formats specifically targeting the medically relevant snake venom toxins [6 , 8] . In the production of serum-based antivenoms , the therapeutically relevant antibodies targeting snake venom toxins cannot easily be separated from the therapeutically irrelevant antibodies targeting other targets ( e . g . bacteria or vira that the immunized animal has encountered during its life . In contrast , recombinant antivenoms may be produced with a significantly higher concentration of therapeutically active antibodies than current serum-based antivenoms , which are known to only contain between 5–36% specific antibodies directed against venom components [9–11] . However , lack of cost-competitive production of antivenom antibody mixtures remains a critical hurdle against making such medicines widely available in poor rural regions of the developing world . Four families of venomous snakes exist ( Elapidae , Viperidae , Atractaspididae , and Colubridae ) , of which the elapids ( such as mambas , cobras , and coral snakes ) and viperids ( such as rattlesnakes and other vipers ) are responsible for the vast majority of envenomings [12] . Generally , viperid venoms are cytotoxic , hemotoxic , and occasionally myotoxic , whereas elapid venoms primarily cause systemic neurotoxicity [12] . The difference in clinical manifestations of viper and elapid venoms stem from the different families of toxins in the snake venom . Further , some of the venom toxins act independently of each other , whereas for others the toxicity is potentiated via toxin synergism [13] . Neurotoxins must first pass the systemic circulation before reaching the relevant targets in the central nervous system and are therefore typically rather small in size . In contrast , toxins which induce tissue damage , including proteases , cytotoxins , and myotoxins , are larger proteins which primarily exert their destructive effects at the site of the bite . This difference in site of action for different toxins means that antivenoms against locally-acting toxins need to be able to reach distal sites and deep tissue [14] , whereas rapid distribution in the circulatory system may be sufficient for effective delivery of antivenoms against systemic toxicity . Currently , animal-derived snakebite antivenoms are manufactured in three different structural formats: IgG-based , F ( ab ) 2-based , and Fab-based [6 , 14] ( see Fig 1 ) . F ( ab ) 2 and Fab-based are generated with the intention of creating improved safety profiles through the removal of the Fc region of the animal-derived IgG antibodies by treatment with pepsin or papain during the manufacturing process [7] . Fab-based antivenoms have been shown to better reach and neutralize toxins in deep tissue than IgG-based antivenoms , but it comes at a cost of reduced serum half-life [14] . In comparison , F ( ab ) 2-based antivenoms have pharmacokinetic properties somewhat in between IgG-based antivenoms and Fab-based antivenoms [14] . Therefore , different antivenom formats have different benefits and drawbacks . The IgG format may represent a more optimal solution for targeting systemically acting toxins , whereas the Fab format may potentially be more ideal against locally acting toxins . Other formats have been investigated , including single-chain variable fragments ( scFv ) and camelid single-domain antibodies ( VHH fragments ) [5–7] , but none of these are have yet been tested in the clinic . The standard method for purification of IgGs from the plasma of immunized animals is based on caprylic ( octanoic ) acid precipitation [15] . This method is inexpensive and robust , but also leaves unwanted traces of caprylic acid in the final product . In order to increase the purity of antivenoms , additional chromatographic steps can be applied , although this leads to lower production yield and significantly higher production cost [16] . Caprylic acid precipitation of IgGs from plasma should be compared to recombinant mAb production , where the standard methods of purification are based on affinity chromatography using protein A capture . The advantage of protein A chromatography is that it leads to a purer product . However , the method is also significantly more expensive , and the resins used for affinity binding of antibodies have a limited number of uses [17 , 18] , typically 150–200 ( personal communication with Anne Tolstrup , Biogen ( Denmark ) A/S ) . For large-scale production of monoclonal recombinant antibodies , Chinese Hamster Ovary ( CHO ) cell-based expression systems are most common [19] , partially due to their ability to produce glycosylation patterns similar to human patterns . CHO cell expression is traditionally performed in a fed-batch process ( see Fig 2 ) , where nutrients for the CHO cells are supplied for a complete manufacturing process followed by harvest of the entire batch . The antibodies from this batch are then purified by one or more chromatographic purification steps . Other cost-competitive production processes , such as hybrid and continuous perfusion processes ( see Fig 2 ) are emerging [20] . In the hybrid process , cultivation is performed in a fed-batch reactor followed by continuous or semi-continuous purification of the produced antibodies . In the continuous perfusion process , the cultivated cells are retained while the growth medium containing the antibodies is continuously substituted with fresh medium in a perfusion reactor . The used medium undergoes a continuous or semi-continuous purification process in order to isolate the antibodies [20 , 21] ( see Fig 2 ) . In the continuous perfusion process , several companies in the industry are employing continuous chromatographic procedures , such as Simulated Moving Bed Chromatography ( SMBC ) ( personal communication with Mads Laustsen , Symphogen A/S ) . This procedure allows for efficient use of chromatographic media as well as it yields higher output of purified product [22] . In the work presented here , we base our calculations on hybrid and continuous perfusion processes using SMBC as downstream process [21] . In addition , we compare the results of these calculations with a manufacturing setup , where the hybrid and continuous processes employ caprylic acid precipitation as purification method [20] , since this downstream process poses as a cost-competitive alternative that is already employed in existing antivenom manufacture . In later years , a growing interest in producing mixtures of monoclonal antibodies has emerged based on an expectation that improved antibody-based therapies can be obtained by targeting more than one target in multiple different indications . Two different strategies can be employed to produce such antibody mixtures: 1 ) Mixing of individually produced parallel batches of each monoclonal antibody ( Fig 3A ) [23–25] and 2 ) oligoclonal expression of antibodies in a single batch ( Fig 3B ) , exemplified by Merus’ Oligoclonics technology [26] and Symphogen’s Sympress technology [23 , 25] . In the Oligoclonics technology , a single cell line is transfected with a shared light chain and 2–3 different heavy chains giving rise to several different specificities . In the Sympress technology , a range of clonal cell lines each expressing different antibodies are mixed together in a single batch in order to produce a mix of a desired number of specific antibodies . The Oligoclonics technology limits the number of different specificities as the single cell line expression system gives rise to stability challenges with higher numbers of different heavy chains [27] . The Sympress technology on the other hand allows for any number of different specificities as each antibody specificity is produced by a separate cell line [23 , 25] . A single monoclonal antibody is unlikely to be able to effectively neutralize a snake venom , given the complexity of snake venoms , which each contain numerous different toxins from different protein families [8] . Instead , mixtures of antibodies will be needed in order to neutralize most of the medically relevant snake venom toxins [6–8 , 28] . Moreover , it is desirable to have polyvalent antivenoms that target the venom toxins from more than just a single snake species , since it is not always possible for clinicians to identify the perpetrating snake species with certainty . Therefore , a clear need exists for technologies enabling cost-competitive production of oligoclonal antibody mixtures , if recombinant antivenoms are to be introduced . Actual proof of concept studies on the production of recombinant antivenoms have never been reported . However , sufficient knowledge from related areas within antibody expression and CHO cell cultivation exists to assess the utility and cost-competitiveness of future recombinant antivenoms . Here , we provide an estimate of the production costs of oligoclonal recombinant antivenoms based on industry data and discuss the implications that novel antibody expression technologies may have on future snakebite envenoming therapy . Antivenoms are likely to be produced with the aim of being able to neutralize most of the medically relevant snakebites in a particular region of the world , but not the entire world . Therefore , we base our present estimates on a hypothetical polyvalent antivenom that could be used to treat the approximately one million snakebite envenomings occurring each year in sub-Saharan Africa , as such pan-African antivenoms derived from animal serum already exist for comparison .
The estimate of the minimum need for active antibodies ( active pharmaceutical ingredient , API ) to neutralize snakebite envenoming is critical for establishing the maximum cost of production for a recombinant antivenom , as this cost is highly dependent on the scale of production . Here , the minimum need for active antibodies is based on the number of snakebite envenomings in sub-Saharan Africa ( approximately one million victims ) [2] , the average number of vials needed to treat an envenoming ( for the representative SAIMR Snake Polyvalent Antivenom from the South African Vaccine Producers ) , which is estimated to be between 6 and 10 vials per treatment [29] , the volume and protein content of this antivenom ( 10 mL and 172 mg/mL ) ( as measured by their absorbance at 280 nm on a NanoDrop 2000c instrument , Thermo Scientific ) [30] , and the lowest estimate ( 5% ) of the percentage of venom-recognizing antibodies in an antivenom ( ranging between 5% and 36% ) [9–11] . In fact , the amount of therapeutically active antibodies is likely to be even lower , since not all venom components are medically relevant , yet still induce an immune response during the animal immunization process . Also , we assume that antibodies produced recombinantly are equipotent ( mol-to-mol ) to the therapeutically active antibodies found in current antivenoms , and we take the difference in molecular weight between IgG molecules and F ( ab ) 2 fragments into account in the calculation of the minimum need of antibodies . Finally , we do not take into account any amount of antivenom that may be discarded due to expiration . The cost of recombinant monoclonal antibodies is highly dependent on the manufacturing approach and the scale of production . The Cost of Goods Manufactured of Active Pharmaceutical Ingredient ( COGMAPI ) has been analyzed extensively on several occasions [20 , 31 , 32] . Production costs of monoclonal antibodies range from 36 USD/g to 1500 USD/g depending on the manufacturing facilities used [20 , 33] and when the process was developed ( personal communication with Anne Tolstrup , Biogen ( Denmark ) A/S ) . We base our calculations on the Sympress technology , as this has been successfully employed in the development and production of several oligoclonal antibody-based therapies that have entered clinical trials ( http://www . symphogen . com/pipeline ) . The COGMAPI of recombinant oligoclonal antibodies is estimated to be 0–10% higher than the production cost of conventional mAbs [25] . In our analysis , four manufacturing situations ( annual production of 100 kgAPI , 200 kgAPI , 500 kgAPI , and 1 , 000 kgAPI ) , based on reported cost estimates for fed-batch , hybrid , and continuous perfusion processes [31] were employed to estimate COGMAPI for oligoclonal antibody production . The reported cost estimates in [31] are calculated based on SMBC as the purification process , using the most realistic assumptions for fed-batch , hybrid , and continuous perfusion , where the fed-batch reactors are used multiple times to obtain titers of 3 g/L , and where continuous perfusion is performed with a cell specific perfusion rate ( CSPR ) of 0 . 05 nL per cell-1 d-1 with a titer of 0 . 4 g/L [21] . To estimate the COGMAPI for a snakebite treatment using oligoclonal recombinant antivenoms at a scale of 500 kgAPI , direct costs of 62 USD/g ( fed-batch ) , 47 USD/g ( hybrid ) , and 89 USD/g ( continuous perfusion ) ( determined in the previous section ) were employed ( also see Fig 4 ) . Estimates were based on two different relative amounts ( 5% and 20% ) of therapeutically active antibodies present in current antivenoms . The number of vials per average treatment for the SAIMR Snake Polyvalent Antivenom from the South African Vaccine Producers was based on data published in [29] . Vial volume and protein concentration are equal to those given above . Finally , to estimate the Cost of Goods Manufactured for the final drug product ( COGMFDP ) the cost of formulation and packaging was ( in discussion with Mads Laustsen , Symphogen A/S ) estimated to be 5 USD/vial , which is five times the cost reported for current antivenoms in [34] , and the upper limit for the concentration of active antibodies in 10 mL vials with recombinant oligoclonal antivenom was set to 100 mg/mL . For comparison , we performed cost estimates based on an alternative set of numbers for production of oligoclonal recombinant antivenom using caprylic acid precipitation as the purification method for hybrid and continuous perfusion processes at a scale of 500 kgAPI . Here , we applied COGMAPI from [20] for commercial production and added 10% for oligoclonal production [25]: 46 USD/g for fed-batch cultivation followed by single-batch chromatography , 33 USD/g for hybrid ( fed-batch cultivation followed by continuous precipitation-based purification ) , and 42 USD/g for continuous perfusion followed by continuous precipitation-based purification .
The unit cost of recombinant antibody production is highly dependent on the scale of antibodies produced [31 , 35] . To estimate the order of magnitude of the amount of antibodies needed to treat all snakebites in sub-Saharan Africa , we disregarded the fact that many antivenoms may not be used for effective therapeutic treatment of victims due to expiration or loss during the supply chain . Thus , we made the assumption that manufacturing volume equals need , which may not be the case in reality . Also , we only considered the therapeutically active antibodies in current antivenoms , which are likely to be no more than the amount of antibodies in antivenoms that are able to recognize snake toxins: 5% to 36% [9–11] , since there is no reason to include ineffective antibodies , when producing antibodies using a recombinant approach . In our assessment of the sub-Saharan African need , we employed the lowest average number of vials ( 6 vials , 10 mL/vial ) reported for the SAIMR Snake Polyvalent Antivenom from the South African Vaccine Producers [29] and the lowest estimate of the content of therapeutically active antibodies in this antivenom , being 5% . Based on these assumptions , we provide the conservative estimate that the sub-Saharan African need for therapeutically active antivenom antibodies to be at least 500 kgAPI per year . This number does not represent a precise estimation of the actual need , however , it provides a lower estimate of the magnitude required for production of antibodies for recombinant antivenoms , which may further be used to estimate the COGMAPI ( USD/gram ) for the manufacture of recombinant antibodies . In reality , the real need may be in the order of 1 to 2 tonAPI per year ( if e . g . the assumed content of therapeutically active antibodies is assumed to be 10–20% instead of only 5% for the SAIMR Snake Polyvalent Antivenom from the South African Vaccine Producers ) . However , increasing the production volumes above 250 kgAPI per year has limited influence on COGMAPI [31] ( disregarding any possible capital investment needed in reality to modify production facilities ) , see Fig 4 . In our assessment of the COGMAPI for oligoclonal recombinant antivenoms produced via a fed-batch , hybrid , and continuous perfusion process , we employed cost estimates provided by industry [31] according to a scale of production in the range of 500 kgAPI per year is estimated in the section above ( see Fig 4 for cost as a function of scale of production ) . At this scale , the COGMAPI for production of oligoclonal antibody mixtures is estimated to 62 USD/g for a fed-batch process , 47 USD/g for a hybrid process , and 89 USD/g for a continuous perfusion process in agreement with similar cost estimates reported in literature [20 , 25 , 31 , 36 , 37] . To compare the COGMAPI per treatment , we here employed both a low estimate ( 5% of the antivenom antibodies are active ) and an estimate that we consider more realistic ( 20% of the antivenom antibodies are active ) based on numbers from [9–11] . In these calculations , we assumed that the average snakebite envenoming requires 8 vials ( 10 mL/vial ) of SAIMR Snake Polyvalent Antivenom [29] . As it can be seen in Fig 5A , the COGMAPI per treatment for oligoclonal recombinant antivenoms at a scale of 500 kgAPI estimated to be between USD 58–233 in fed-batch mode , USD 44–176 in hybrid mode , and USD 83–334 in continuous perfusion mode based on the low and high estimates of the content of therapeutically active antibodies in an antivenom . In order to estimate the Cost of Goods Manufactured for the Final Drug Product ( COGMFDP ) , which includes formulation and packaging , we used a cost estimate of 5 USD/vial . We further assumed that each 10 mL vial could be formulated with an antibody concentration of up to 100 mg/mL , but that for administration purposes an average envenoming should be treated with at least 4 vials . This would allow a physician to easily administer less antivenom for milder envenomings requiring a dose lower than average . Since much fewer antibodies are needed in a recombinant antivenom compared to traditional antisera , as only the therapeutically active antibodies are manufactured , these criteria had the effect that treatments with all antibody formats would meet the minimum requirement of 4 vials for formulation , adding 20 USD to the COGM of treatment . The COGMFDP per treatment is therefore estimated to USD 78–253 in fed-batch mode , USD 64–196 in hybrid mode , and USD 103–354 in continuous perfusion mode ( see Fig 5A ) . The fairly broad estimates for each individual situation derives from the uncertainty of how many therapeutically active antibodies actually exist in the existing SAIMR Snake Polyvalent Antivenom . In these estimates , we employed a percentage between 5% and 20% , which corresponds to a 4-fold increase from the lowest estimate to the highest . More knowledge and better understanding of the therapeutic content of existing antivenoms would be useful to narrow down the cost estimates presented here . Even more desirable , however , would be to have proof of concept recombinant antivenoms with a defined set of discovered and tested human monoclonal antibodies upon which better cost estimates could be performed . As effective doses of individual antibodies in real case scenarios may span 1–2 log units , having such a defined set of antibodies with specific ED50s would allow for bottom-up calculations based on optimal molecular ratios between antibodies ( removing the need for assuming equipotency ( mol-to-mol ) of antiserum-based antibodies and recombinantly produced antibodies , see Methods ) . In real case scenarios , minimum requirements for antibody efficacy and ease of expression should be imposed as selection criteria to ensure that the antibodies selected for a recombinant antivenom are compatible with oligoclonal expression methods ( e . g . Sympress or Oligoclonics ) . Since caprylic acid precipitation is routinely employed in the manufacture of current antivenoms [15] , a comparison of the COGMFDP per treatment for oligoclonal recombinant antivenoms produced using continuous caprylic acid precipitation for the hybrid and continuous perfusion processes was performed . Based on numbers from [20 , 25] , the COGMAPI was in this case estimated to 46 USD/g for fed-batch cultivation followed by single-batch chromatography , 33 USD/g for hybrid ( fed-batch cultivation followed by continuous precipitation-based purification ) , and 42 USD/g for continuous perfusion followed by continuous precipitation-based purification . These cost estimates based on the application of caprylic acid precipitation for the hybrid and continuous process are significantly lower than the estimates based on processes employing chromatography . Using these lower cost estimates therefore leads to a significant lowering of the COGMAPI , estimated to USD 43–173 for fed-batch mode , USD 31–124 for hybrid mode , and USD 39–158 for continuous perfusion mode . When including 4 vials for formulation , the COGMFDP is estimated to USD 63–193 for fed-batch mode , USD 51–144 for hybrid mode , and USD 59–178 for continuous perfusion mode ( see Fig 5B ) . Our results may therefore point towards the applicability of caprylic acid precipitation for manufacture of recombinant antivenoms due to its lower cost and compliance with current antivenom manufacture . In general , since we estimate the cost estimates on oligoclonal expression by adding 10% to the COGMAPI for oligoclonal antibodies compared to monoclonal antibodies , the cost estimates presented here subtracted these 10% may also be valid for the production of recombinant antivenoms based on monoclonal antibodies against animal venoms , where venom toxicity can be abrogated by targeting a single toxin ( or more toxins with a cross-reactive antibody ) . The COGM estimates ( approx . USD 60–350 per treatment for processes using chromatography as purification method and approx . USD 50–190 per treatment for processes using caprylic acid precipitation as purification method ) presented here compare favorably with the wholesale cost of SAIMR Snake Polyvalent Antivenom ( USD 640 per treatment ) [29] . Although the wholesale cost may not adequately reflect the COGMFDP for SAIMR Snake Polyvalent Antivenom , our COGMFDP estimates for recombinant production indicate that even with a fairly high margin for distribution , sales , and profit ( 45–92% depending on manufacturing strategy and purification method ) , that snakebite envenoming treatments based on recombinant antivenoms could be cost-competitive with current serum-based antivenoms . In this feasibility study , we focus on the production of IgG-based antibodies , since other antibody formats , such as Fab , scFv , or VHH would not be produced in CHO cells , but rather in prokaryotic systems ( personal communication with Mads Laustsen , Symphogen A/S ) . We therefore do not intend to provide an in-depth discussion on any therapeutic or pharmacokinetic benefits that some antibody formats may have above others . Some of these differences include the bivalency of IgG and F ( ab ) 2 and the prolonged half-life of IgG compared to both F ( ab ) 2 , Fab , scFv , and VHH fragments [6 , 14 , 38 , 39] . The shorter half-life of F ( ab ) 2s , Fabs , scFvs , and VHHs makes it likely that larger doses of these more rapidly degraded formats are needed to effectively neutralize all venom toxins in a snakebite envenoming . Also , even for systemically acting neurotoxins that are able to reach their targets within minutes , a delayed release of toxins from the bite site may occur over a period of days , demanding either repeated doses or longer serum half-life of the antivenom [40] . On the other hand , smaller fragments may possibly penetrate faster into distal tissue , where venom toxins are initially located following a snakebite , possibly providing therapeutic benefits to the use of smaller fragments . It is likely that Fabs , scFv , or VHH-based recombinant antivenoms could be produced in large scale at an even lower cost using prokaryotic systems that do not require expensive media , and which have a high growth rate and productivity . A COGM analysis comparing the different formats and taking the molecular and pharmacokinetic differences into account is , however , beyond the scope of this study as we deem it to be too speculative , given the large amount of assumptions that would need to be made . Since the majority of current antivenoms ( including the SAIMR Snake Polyvalent Antivenom ) are based on the F ( ab ) 2 format [6] , which has a comparable molecular size and bivalence , the COGM estimates presented here may indeed provide a rough assessment of the relative cost-competitiveness of IgG-based antivenoms . Moreover , the human IgG format with its long half-life has never been thoroughly explored in the field of antivenom , and it may indeed provide additional benefits such as lower administration doses needed for the treatment of a snakebite envenoming . The data presented here therefore warrant further research into the development of recombinant antivenoms . Additionally , a need presents itself for discussing how to handle regulatory affairs , when a switch from conventional antisera ( which are defined as “blood products” ) to recombinant antivenoms ( probably to be defined as “biopharmaceuticals/biologics” ) occurs . As a final remark , the cost of recombinant antibody production has been hypothesized to possibly reach as low levels as 10–20 USD/g in the longer term [33 , 36 , 41] . Although reaching such a low cost of production may be years into the future and require a larger scale than what is relevant for antivenoms , it may indicate that the cost of recombinant antivenoms in the future may be even lower than what is estimated here .
Based on industrial cost estimates of oligoclonal expression of antibodies at large scale and the estimated need for therapeutically active antibodies to treat the one million snakebite envenomings occurring each year in sub-Saharan Africa , we estimate that recombinant pan-African antivenoms could be produced with a COGMAPI of 62 USD/g ( fed-batch ) , 47 USD/g ( hybrid process ) , and 89 USD/g ( continuous perfusion ) , when chromatography is used as purification method , and 46 USD/g ( fed-batch ) , 33 USD/g ( hybrid process ) , and 42 USD/g ( continuous perfusion ) , when caprylic acid precipitation is used as purification method for the hybrid and continuous perfusion process . This translates into an approximate Cost of Goods Manufactured for the IgG-based final drug product including vials and formulation ( COGMFDP ) between approx . USD 60–350 per treatment ( for processes using chromatography as purification method ) and approx . USD 50–190 per treatment ( for processes using caprylic acid precipitation as purification method ) for treating an average snakebite envenoming that would normally require 8 vials of SAIMR Snake Polyvalent Antivenom from the South African Vaccine Producers [29] . These estimates compare favorably with previously reported costs of treatment ranging from USD 56 to 640 for current serum-based antivenoms [29] , leaving a large margin ( 45–92% ) for formulation , distribution , sales , and profit when comparing with the golden standard antivenom , SAIMR Snake Polyvalent Antivenom . The data-supported cost estimates presented here may provide better insight into the economics of recombinant antivenom production and help guide decision processes on what technological platforms future antivenoms should be built on . Our estimates demonstrate the cost-competitiveness of the production of recombinant antivenoms , which may provide incentive to more researchers to engage in the development and testing of such novel therapies against snakebite envenoming . | Given the medical importance of snakebite envenoming and the current shortage of antivenoms in sub-Saharan Africa , technological advances in antivenom development and production are needed . One of the avenues that could be taken involves the use of recombinant antivenoms based on oligoclonal mixtures of human IgG antibodies , since these may have the benefits of being compatible with the human immune system and their production is independent on animal immune systems and venom procurement . However , an important aspect of introducing recombinant antivenoms to the clinic is their cost of production given that snakebite victims are often poor rural workers living in remote parts of the tropical parts of the developing world . Here , we aim to provide cost estimates of recombinant antivenom manufacture with special focus on snakebite envenoming in sub-Saharan Africa . Our results indicate that recombinant antivenoms in the future will indeed be cost-competitive compared to existing animal-derived serum-based antivenoms . Furthermore , we outline different manufacturing strategies and suggest the use of caprylic acid precipitation as a low cost purification method following cultivation of CHO cells for antibody expression due to its use in current antivenom manufacture . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion",
"Conclusions"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"toxins",
"pathology",
"and",
"laboratory",
"medicine",
"immunology",
"tropical",
"diseases",
"biological",
"cultures",
"vertebrates",
"animals",
"toxicology",
"toxic",
"agents",
"reptiles",
"neglected",
"... | 2017 | Recombinant snakebite antivenoms: A cost-competitive solution to a neglected tropical disease? |
Skeletal muscle morphogenesis transforms short muscle precursor cells into long , multinucleate myotubes that anchor to tendons via the myotendinous junction ( MTJ ) . In vertebrates , a great deal is known about muscle specification as well as how somitic cells , as a cohort , generate the early myotome . However , the cellular mechanisms that generate long muscle fibers from short cells and the molecular factors that limit elongation are unknown . We show that zebrafish fast muscle fiber morphogenesis consists of three discrete phases: short precursor cells , intercalation/elongation , and boundary capture/myotube formation . In the first phase , cells exhibit randomly directed protrusive activity . The second phase , intercalation/elongation , proceeds via a two-step process: protrusion extension and filling . This repetition of protrusion extension and filling continues until both the anterior and posterior ends of the muscle fiber reach the MTJ . Finally , both ends of the muscle fiber anchor to the MTJ ( boundary capture ) and undergo further morphogenetic changes as they adopt the stereotypical , cylindrical shape of myotubes . We find that the basement membrane protein laminin is required for efficient elongation , proper fiber orientation , and boundary capture . These early muscle defects in the absence of either lamininβ1 or lamininγ1 contrast with later dystrophic phenotypes in lamininα2 mutant embryos , indicating discrete roles for different laminin chains during early muscle development . Surprisingly , genetic mosaic analysis suggests that boundary capture is a cell-autonomous phenomenon . Taken together , our results define three phases of muscle fiber morphogenesis and show that the critical second phase of elongation proceeds by a repetitive process of protrusion extension and protrusion filling . Furthermore , we show that laminin is a novel and critical molecular cue mediating fiber orientation and limiting muscle cell length .
Muscle specification and morphogenesis during early development are critical for normal muscle physiology . In vertebrates , most of the musculature is derived from somites [1]–[3] . Somites are segmentally reiterated structures delineated by somite boundaries . As development proceeds , a portion of the somite gives rise to skeletal muscle fibers that comprise the myotome . The terminal ends of myotomal muscle fibers attach to somite boundaries , which then become myotome boundaries . In teleost fishes , myotome boundaries give rise to the myotendinous junction ( MTJ ) [4] . Myotome development is perhaps best understood in amniotes . Myogenesis in amniotes begins when muscle precursor cells translocate from the overlying dermomyotome to the myotome [3] . The first myocytes to translocate come from the dorsomedial lip , but later in development myocytes translocate from all dermomyotome borders as well as the central region [5]–[7] . Time-lapse analysis in the chick embryo has shown that spatial domains of the somite differ in cell behaviors that generate the primary myotome [8] . The above studies have elucidated cell movements that generate the myotome . However , although it is known that short , mononucleate , muscle precursor cells generate long , functional multinucleate muscle fibers , it is not known how this occurs . Interestingly , the early zebrafish and chick myotomes have been described as containing mononucleate muscle fibers [8] , [9] . Our intent in undertaking this study was to utilize the advantages of the zebrafish system to shed light on early muscle development in a vertebrate model . Elucidation of the cellular mechanisms that underlie muscle fiber development and tendon attachment is critical for a comprehensive understanding of muscle development . Towards this end , muscle morphogenesis has been studied in different model systems . C2C12 myoblasts in culture elongate slightly prior to differentiation , align with each other , and fuse to generate a multinucleate myotube [10] . We have called this scenario elliptical growth . In grasshopper embryos , the first muscle cells elongate between attachment sites prior to fusion . These cells extend many processes in multiple directions while elongating [11] . We have termed this scenario branching . Elegant studies in Drosophila have shown that muscle morphogenesis occurs when myoblasts fuse to generate long , multinucleate myotubes and identified a number of proteins required for myoblast fusion [12] , [13] . Extremely exciting recent studies have shown that there is some conservation of molecular mechanisms that mediate muscle cell fusion between Drosophila and zebrafish [14] , [15] . Interestingly , however , the primary myotome in chick and zebrafish is mononucleate [8] , [9] . This suggests that myoblast fusion does not mediate the earliest stages of muscle morphogenesis in vertebrates , but occurs after initial muscle fiber elongation . These distinct mechanisms of morphogenesis in different systems highlight the fact that a mechanistic study of muscle fiber morphogenesis in vertebrates has not yet been undertaken . Identification of discrete morphogenetic steps that mediate muscle fiber morphogenesis in vertebrates is necessary to provide a framework for future molecular analyses . Adhesion of muscle fibers to the basement membrane is critical for muscle function . The basement membrane attaches muscle fibers to connective tissue that then attaches to the skeletal system; this attachment is critical for force transduction from muscle to bone . One major component of the basement membrane is laminin . Laminin is a heterotrimeric protein composed of α , β and γ subunits that generate at least 15 different isoforms [16] . The importance of laminins in muscle physiology is evidenced by the fact that mutations in lamα2 result in muscular dystrophies [17]–[20] . Recent work has shown that muscle fibers in zebrafish mutant for lamα2 elongate and attach to the MTJ , but at 48 hours post fertilization ( hpf ) fibers detach before death , providing novel insight into roles for lamα2 in muscle disease [21] . Significantly less is known about spatiotemporal mechanisms of basement membrane assembly during early skeletal muscle development and whether adhesion to the basement membrane contributes to morphogenesis . Recent data suggest that the laminin receptor Integrin α6β1 is necessary for both basement membrane assembly and normal expression of myogenic regulatory factors in cultured mouse explants [22] . However , Integrin α6β1 binds to multiple laminins with distinct affinities [23] and roles for individual laminin chains during early muscle development in vivo have not been identified . In order to determine whether the basement membrane is critical during early development for muscle fiber elongation and attachment , it is first necessary to understand the cellular basis of muscle fiber elongation and attachment . The relative simplicity of zebrafish skeletal muscle , where slow and fast-twitch fibers are spatially segregated , makes it an ideal model system to study muscle cell elongation and MTJ morphogenesis . Morphogenesis of the somite boundary into the MTJ involves three stages: initial epithelial somite boundary formation , transition and myotome boundary/MTJ formation [24] . Transition encompasses the lateral displacement of slow-twitch muscle fibers and the subsequent elongation and differentiation of fast-twitch fibers [25]–[28] . The initial myotome forms by 26 hpf and contains long muscle fibers attached to the myotome boundary/MTJ . At this point , the extracellular matrix ( ECM ) proteins Fibronectin , laminin and Periostin concentrate at the MTJ [29]–[31] . Morpholino-mediated inhibition of Periostin disrupts MTJ formation [29] , [32] , [33] , but discrete and mechanistic requirements for other ECM proteins and their receptors are not known . In addition , the precise mechanism by which elongating muscle fibers attach to the MTJ and cease elongation has not been elucidated . The purpose of this study was to rigorously and quantitatively characterize , for the first time in vertebrate embryos , the cellular events that generate long myotubes from initially short muscle precursor cells . We focused on fast-twitch fiber morphogenesis in zebrafish embryos . Our goal was to develop methods with which discrete functions for proteins involved in muscle morphogenesis could be identified . Towards this goal , we utilized time-lapse analysis , genetic mosaic analysis , and three different mathematical tools including a powerful wavelet-based image analysis formalism to provide novel insight into cellular and molecular mechanisms that underlie muscle fiber elongation and subsequent attachment to the nascent MTJ .
Although the elongation of somitic cells is critical for actin-mediated contractility that underlies muscle function , the cellular and molecular basis of elongation in vertebrates is not well understood . An understanding of how muscle cells elongate is critical in order to determine mechanistic roles for genes required in elongation . We used time-lapse microscopy of zebrafish embryos labeled with BODIPY-Ceramide to outline cells . This type of time-lapse analysis , where all cells are labeled , provides an initial framework with which to focus further investigation into fast-twitch fiber morphogenesis . Fast-twitch muscle cells can be identified because , in contrast to slow-twitch muscle fibers , they are not migrating medially-laterally [25] . The transition from a somite to a myotome is a dynamic process ( Figure 1 , Movie S1 ) with at least three phases . The first phase is short muscle precursor cells . Second , muscle fibers elongate by extending narrow protrusions to intercalate between other cells ( Figure 1 A2 at 80 min , blue pseudocolored cell ) . Elongation ends when cells adhere to the anterior and posterior boundaries . The third phase is myotube formation . Recently elongated cells are long , but irregularly shaped ( Figure 1 A1 green cell at 0 min , A2 blue cell at 84–168 min , Movie S1 ) . During myotube formation , long cells with grooves continue to change shape until they form a more uniformly shaped tube without grooves ( Figure 1 A1 green cell at 208 min ) . An additional time-lapse is shown in Movie S2 . A three-dimensional quantification of cell morphology is critical to distinguish between scenarios of muscle cell elongation ( Figure 2A ) . We transplanted dextran-filled cells into unlabeled host embryos and three-dimensionally reconstructed the behavior of labeled cells through time ( Figure 2B ) . For each time point , the z-series was three-dimensionally projected and the area , perimeter , and major axis were measured . Thus two-dimensional parameters ( area , perimeter , and major axis ) were obtained from three-dimensional projections of cells . The analysis of labeled cells in an unlabeled field of cells allows unambiguous determination of cellular shape dynamics and quantification of morphometric parameters . We analyzed cell behaviors in two ways: ( 1 ) analysis of the filament index and ( 2 ) analysis of the relative dynamics of area and perimeter changes through time . As shown below , this approach supports and extends what was observed in BODIPY-Ceramide labeled embryos . The filament index is an excellent mathematical parameter that describes cell morphology . The filament index is a measure that quantifies the departure of a shape from a circle ( see methods ) . A circle has a filament index of 1 and a higher filament index indicates a larger departure from a circular shape . Short muscle precursor cells have a low filament index ( FI ) indicating that their morphology is close to a circle ( Figure 2F , G1 , FI = 1 . 6±0 . 6 , Table 1 ) . Short muscle precursor cells extend and retract very short ( <2 µm ) filopodia-like protrusions in all directions ( Figure 2C , Movie S3 ) . Small changes in the area , perimeter , and length of muscle precursor cells reflect the dynamic shape changes of precursor cells ( not shown ) . However , their overall shape and size remains consistent . Elongating cells lengthen towards their attachment site , the MTJ . Elongating cells have a higher filament index than short precursor cells ( Figure 2F , 2 . 9±0 . 8 , Table 1 ) . The filament indices of elongating cells increase slightly through time ( Figure 2G1 , G2 ) , reflecting their departure from a circular shape . One purpose of this experiment was to distinguish between possible scenarios of muscle fiber elongation summarized in Figure 2A . The difference between the fusion and remaining scenarios is the timing of fusion relative to elongation . In the fusion scenario , fusion of short myoblasts is the major morphogenetic event that drives fiber elongation . Fusion of multiple short cells generates a long , multinucleate myotube in one step as in Drosophila [12] . In the remaining scenarios , cells elongate prior to fusion . We analyzed nuclear content of elongating and recently elongated cells and found that mononucleate fast-twitch cells elongate to the MTJ prior to fusion ( Figure 4F–H , n = 108 cells ) . Thus , the first fast-twitch fibers in zebrafish do not fuse prior to elongation . The remaining scenarios are branching , elliptical growth , and protrusion ( Figure 2A ) . The difference between the branching scenario and the elliptical growth/protrusion scenarios is the amount , size , and direction of protrusive activity . In grasshopper , the first muscle cells to elongate have extensive protrusions in many different directions [11] . This is depicted in the branching scenario ( Figure 2A ) . In contrast , the elliptical growth and protrusion scenarios depict cells that elongate in a fixed direction . Both time-lapse analysis and analysis of cell morphology in fixed embryos indicate that fast-twitch fibers in zebrafish embryos elongate in a fixed direction and do not exhibit a branching morphology with multiple protrusions extended in different directions ( Figures 2 , 4 ) . Rather , fast-twitch cells extend long ( >4 µm ) protrusions along their long axis ( Figure 2D , Movie S3 ) . These results indicate that the branching scenario does not apply to initial fast-twitch muscle morphogenesis in zebrafish . The two remaining scenarios differ in the nature of elongation . The elliptical growth scenario depicts elongation as a continuous process reminiscent of a balloon filling . The protrusion scenario suggests that elongation is incremental and proceeds via a 2-step mechanism: protrusion extension and protrusion thickening . Time-lapse analysis suggests that fast-twitch cells extend protrusions that subsequently thicken ( Figures 1 , 2 ) . Geometrical models were developed and used to determine how area and perimeter would change through time if cells were elongating via the two different scenarios . The major difference between the two models is the nature of dynamic changes in area and perimeter during elongation ( see methods for details ) . Area and perimeter increase linearly in the elliptical growth model ( Figure 3A ) and incrementally in the protrusion model ( Figure 3 A1 ) . The incremental nature of growth in the protrusion model is because the perimeter increases slightly more when the protrusion extends , but the area increases slightly more when the protrusion thickens ( Figure 3 A1 ) . Therefore the difference between the two models is how area and perimeter values change as the cell grows: linear changes occur in the elliptical growth model and incremental changes in the protrusion model . In all cells examined , the rate of area increase is higher than the rate of perimeter increase as is predicted by both models ( Figure 3 B3 ) . However , area and perimeter increase incrementally during fast-twitch muscle cell elongation ( Figure 3 B1 , B2 ) . Thus , analysis of area and perimeter dynamics supports the two-step intercalation model . Boundary capture occurs when elongating muscle cells reach myotome boundaries and stop extending . The filament index of cells in the boundary capture/myotube formation phase is significantly higher than the preceding two phases ( Figure 2F , Table 1 ) . Their filament indices decrease slightly through time ( Figure 2 G1 , G2 ) . This decrease reflects the fact that a rod-shaped cell has a similar perimeter and length , but larger area than a long , irregularly shaped cell . Recently elongated cells can be irregularly shaped and of varying diameters ( Figure 2E ) . The thinner portions of the cell then thicken until the entire cell consists of a more uniform diameter ( Figure 2E , Movie S3 ) . Note that changes in area and perimeter of cells in the myotube formation phase are distinct: in this phase the area increases much more than the perimeter ( Figure 3C , compare C3 to B3 ) . The increase in area without a substantial increase in perimeter reflects the adoption of a more tube-shaped , regular morphology in the myotube formation phase . Taken together , the time-lapse data along with two different quantitative analyses ( area and perimeter dynamics through time and the filament index ) indicates that there are three discrete morphogenetic phases that generate the first fast-twitch muscle fibers: short muscle precursor , intercalation/elongation and boundary capture/myotube formation . Morphometric analysis of fixed cells corroborates the time-lapse data . Myotome formation proceeds in an anterior-posterior progression . Thus , this approach allows analysis of muscle cells in various stages of elongation within the same embryo . As observed in live embryos , fixed muscle precursor cells are short ( <5 µm ) and have short protrusions ( Figure 4A , F , white arrowheads ) . Protrusions in fixed cells are also observed to extend in all directions ( Figure 4A ) . The filament index of live precursor cells and fixed precursor cells is similar ( Figure 4E live cells 1 . 6±0 . 6 , fixed cells 1 . 4±0 . 2 ) . The correlation of the length with perimeter is also similar to live cells ( Table 1 ) . Both the qualitative appearance and the morphometric properties of fixed cells presumed to have been elongating ( those between 5 µm and 40 µm in fixed embryos ) are similar to live elongating cells . The major axis is very strongly correlated with perimeter and area in both populations ( Table 1 ) . Similar to live cells , long narrow protrusions are only observed along the major axis ( Figure 4B , G , yellow arrowheads , Movie S4 ) . We also analyzed the nuclear content of dextran-filled cells during elongation . A z-series was taken and cells were examined in three dimensions . No elongating cells contained more than one nucleus ( Figure 4G , n>100 cells ) . Cells that were fully elongated but irregularly shaped were presumed to be in the boundary capture/myotube formation phase ( Figure 4C is a three-dimensional projection of an irregularly shaped but elongated cell ) . The filament index of these cells was also similar to live cells ( Figure 4E ) . Fast-twitch cells in this phase were mononucleate ( Figure 4 H shows one focal plane of a mononucleate cell that is elongated but irregularly shaped when examined in three dimensions , Movie S4 , n = 108 ) . Interestingly , all muscle cells that contained multiple nuclei exhibited a stereotypical tubular shape ( n = 57 , see Figure 4I ) . These data suggest the intriguing possibility that the transition from an irregularly shaped long cell to a rod-shaped myotube may involve fusion . Our use of dextran-labeled cells in a field of unlabeled cells clearly highlights the morphological complexity of elongating fast-twitch muscle cells and indicates that it is not possible to unambiguously identify multinucleate cells utilizing a nuclear marker as well as a marker that denotes all cells ( such as phalloidin ) . Thus , we do not know the exact timing of muscle cell fusion or whether fusion contributes to the morphogenesis of irregularly shaped long fibers into regularly shaped , cylindrical myotubes . However , it is evident that the first fast-twitch muscle cells do not fuse in order to elongate . The filament index of fixed muscle cells , as in live cells , is significantly different between each phase ( data not shown , two-tailed t-test , p<0 . 01 for all comparisons ) . These data support the time-lapse analyses and provide new tools for analysis of morphogenetic defects in various mutant/morphant embryos . The identification of discrete , mathematically distinct phases provides a paradigm by which muscle morphogenesis in mutant embryos can be assessed . The above data also indicate that the morphology of fixed cells is not significantly different than live cells . However , although obtaining single labeled cells within a field of unlabeled cells in fixed embryos is easier than time-lapse analysis , it is not feasible in all model systems . We thus looked for a different mathematical tool to quantify cellular organization . Ideally such a tool would allow objective quantification of cellular structure with an easier experimental preparation such as staining with phalloidin to outline all cells . Therefore , we adapted and applied the 2D Wavelet-Transform Modulus Maxima ( WTMM ) method [34] , [35] . This method can be used to quantify the amount of structure , or order , of objects that do not necessarily have a well defined boundary . We used this approach to quantify the structural organization of cellular lattices during muscle fiber elongation . The WTMM analysis filters an image with the gradient of a smoothing function ( i . e . a wavelet ) at a given size scale . Places within the image where the intensity variation is maximal are given by the wavelet-transform modulus maxima ( i . e . the WTMM ) . Next , the positions of maximal intensity variation along these maxima chains are identified . These are the WTMM maxima , or WTMMM . At these nodes , the direction where the signal has the sharpest variation is calculated . An arrow that points upward has an angle of π/2 and an arrow that points down has an angle of −π/2 . The anisotropy factor Fa is then calculated from the probability density function , Pa ( A ) , of the angles A of the WTMMM vectors . Fa is defined in such a way that randomness , isotropy , has a value of Fa = 0 . Any value of Fa>0 quantifies the extent of departure from isotropy . A randomly structured cell lattice has arrows pointing in all directions and a low anisotropy factor . The arrows point in all directions because the direction of maximal intensity variation is random . A more organized cell lattice will have more arrows pointing in the same direction and a stronger anisotropic signature . More arrows will point in the same direction in an ordered cell lattice because the direction of maximal intensity variation will be the same between multiple cells . Thus , this formalism objectively provides a quantitative assessment of morphological structure . A step-by-step explanatory diagram is presented in Figure 5 . The WTMM analysis was applied for all size scales between a∼4 and a∼13 µm ( see methods for details on staining and image preparation ) . Short muscle precursor cells have a low anisotropy factor indicating that there is only a small departure from isotropy . Organization increases throughout muscle elongation . This increase in organization through time is visible as the increase in the proportion of arrows pointing in the same directions ( Compare Figure 5B to 5E ) . The averaged probability density functions Pa ( A ) over all size scales a are shown in Figure 5F and the resulting averaged anisotropy factors Fa are shown in Figure 5G . Each phase of muscle development has a significantly higher anisotropy factor indicating increasing cellular organization through time ( importantly , statistical significance is maintained when only single size scales are used ) . Taken together , all the methods used ( time-lapse analysis , area/perimeter dynamics , filament index , and 2D WTMM ) show that there are discrete phases of fast muscle morphogenesis . Furthermore , the fact that the 2D WTMM analysis supports the other morphometric analyses used indicates that this is an exceedingly valuable tool that can objectively quantify how ordered/structured a field of cells is without having to isolate or segment individual cells . Thorough knowledge of the cellular mechanisms underlying muscle fiber elongation provides a framework for elucidating the molecular basis of muscle cell elongation . We asked if a prominent basement membrane protein , laminin , is required for muscle morphogenesis . It is known that a laminin receptor , Integrin α6β1 , is required for normal myofiber development in cultured mouse explants [22] but the relevant laminin ligands are unknown . We find that muscle cell elongation in lamβ1 and γ1 mutants and morphants is delayed . In zebrafish , slow-twitch fibers migrate laterally and trigger fast muscle cell elongation [28] . Thus , slow fiber location is an excellent marker for assaying fast muscle cell elongation: fast cells medial to slow fibers should be fully elongated . Although slow muscle fiber migration is disrupted in lamβ1 and γ1-deficient embryos ( Figure S1 ) , some slow fibers migrate laterally . However , fast muscle cells medial to migrating slow fibers are short in lamβ1 or lamγ1 mutants and morphants ( Figure 6B , C , and data not shown , n = 6 grumpy/lamβ1 mutant embryos , 16 lamβ1 morphant embryos , 5 wi390/lamγ1 mutant embryos and 10 lamγ1 morphant embryos ) . Fast-twitch muscle cells belatedly elongate in lamβ1 and γ1-deficient embryos and the filament index of cells in all three phases is similar to control embryos ( Table 1 ) . However , fast muscle cells frequently appear misoriented in lamβ1 and γ1-deficient embryos ( Figure 6E , note the abnormal angle of cells that are not aligning in a parallel array , data not shown ) . Application of the 2D WTMM method indicates that myotubes in lamγ1-deficient embryos are significantly more disorganized than in control embryos . Elongated myotubes in control embryos form an organized array as indicated by the strong polarization of the yellow arrows ( Figure 6 G3 ) . The arrows tend to point either up or down resulting in high peaks at π/2 and −π/2 ( Figure 6K lime green line ) and a higher anisotropy factor ( Figure 6L ) . In contrast , arrows in lamγ1-deficient embryos are far less polarized ( compare Figure 6 H3 to G3 ) . The peaks at π/2 and −π/2 are lower than in wild-type embryos ( Figure 6K lime green line ) and the anisotropy factor is significantly lower ( Figure 6L ) . Thus , application of the 2D WTMM formalism quantitatively supports the qualitative perception that muscle fibers are disorganized in laminin-deficient embryos . The next question that follows is when does the anisotropic signature in laminin-deficient embryos become different from wild-type embryos ? No overt morphological differences between control and laminin-deficient cells in the short precursor phase are visible to the eye ( Figure 6I , J ) . However , there is a slight but significant difference between the anisotropy factors ( Figure 6L ) . The difference between anisotropy factors increases at every phase of muscle morphogenesis . These data indicate that laminin is required for cellular organization as early as the short precursor phase . Thus , subsequent myotube disorganization may reflect both early and late requirements for laminin during muscle morphogenesis . It has been proposed that muscle fibers elongate until they reach a small patch of ECM that functions to capture elongating muscle cells and prevent them from extending into the next myotome [24] . However , it is not known which of the many ECM components of the MTJ are required or if multiple proteins are required . We find that both lamβ1 and lamγ1 play a role in MTJ morphogenesis . Some fast muscle cells in lamβ1 and lamγ1 mutants and morphants do not stop elongating at the MTJ ( Figure 6 F1 white arrowhead , note that the muscle cell extends a long , thin protrusion across the boundary ) . The MTJ is visible in 48 hpf wild-type ( WT ) embryos as a dark line devoid of filamentous actin ( Figure 7A , white arrow ) . In wi390/lamγ1 mutant embryos , some muscle fibers inappropriately cross the MTJ and are approximately twice as long as their counterparts that did not cross the boundary ( Figure 7 A1 red arrowhead ) . These cells are multinucleate ( data not shown ) , indicating that boundary capture is not required for fusion . The crossing of a boundary by a few muscle fibers results in an asymmetrical myotome: some of the myotome has longer fibers while the majority of fibers are an appropriate length ( Figure 7 A1 ) . Fast fibers in gup/lamβ1 mutants also cross MTJ boundaries ( Figure 7 A2 red arrowhead ) . At 48 hpf , some boundaries were crossed within every laminin-deficient embryo examined . Generally , 16–24% of boundaries were crossed ( average % of boundaries crossed: WT , 0% , n>100; gup/lamβ1 , 20% crossed , n = 12 embryos , 3 experiments; wi390/lamγ1 , 22% crossed , n = 9 embryos , 1 experiment; lamβ1 MO , 24% crossed , n = 26 embryos , 5 experiments; lamγ1 MO , 16% crossed , n = 18 embryos , 3 experiments ) . Our data show that lamβ1 and γ1 play a role in boundary capture of elongating muscle fibers , but the mechanism of capture is not yet known . A dense network of polymerized laminin may function as a physical barrier that stops elongating muscle fibers . Interestingly , however , laminin polymerization can trigger changes in the organization of the matrix , ECM receptors and cytoskeletal components [36] . Cell-autonomous changes in cytoskeletal organization upon laminin binding provide an alternate hypothesis: that signaling that results from laminin binding may mediate boundary capture in a cell-autonomous fashion . We hypothesized that WT cells transplanted in laminin-deficient embryos might be able to secrete small amounts of laminin that would facilitate their capture and reduce the likelihood of elongating through the boundary . To test this , cells from dextran-injected control embryos were transplanted into lamγ1 morphant hosts . Control cells were less likely than lamγ1 morphant cells to cross the boundary ( Figure 7C , D ) . Only 6 percent of control cells crossed boundaries ( 19/311 cells ) whereas 25% of morphant cells crossed boundaries ( 407/1631 cells ) . Control cells undergo boundary capture even when adjacent to lamγ1 morphant cells crossing boundaries ( Figure 7 B1 , B2 note that the red control cell , white arrowhead respects the boundary , but adjacent morphant cells cross the boundary , red arrowhead ) . These data not only provide the first evidence that laminin plays a role in ceasing initial myofiber elongation , but the cell autonomous rescue of boundary integrity by WT cells suggests that boundary capture is mediated at the single cell level .
Both qualitative and quantitative assessments of early muscle development are critical to facilitate identification of molecular mechanisms that underlie morphogenesis . We find that the three phases of early fast muscle morphogenesis are qualitatively and quantitatively different . These stages are short muscle precursor cells , elongating muscle cells and myotube formation . Short muscle precursor cells have a low filament index and extend and retract short ( <2 µm ) protrusions in all directions . Elongating fast muscle cells extend long protrusions along the axis of elongation and have a higher filament index . Long muscle cells forming myotubes have an even higher filament index indicating yet a further departure from a circular shape . Thus , we provide a novel paradigm whereby morphometric analysis can distinguish different phases of early muscle development . It is not known how the first fast-twitch muscle cells elongate during vertebrate development . We utilized an experimental approach to distinguish between potential scenarios ( Figure 2A ) . C2C12 myoblasts in culture elongate prior to differentiation and fuse to generate a multinucleate myotube [10] that we termed the elliptical growth scenario . The first muscle cells to elongate in grasshopper embryos ( muscle pioneers ) exhibit a morphology similar to that of pathfinding neurons [11] , we have called this the branching scenario . During Drosophila embryogenesis , muscle cells elongate via fusion [12] , [13] and zebrafish homologues of genes required for muscle cell fusion in Drosophila are also required for normal muscle development in zebrafish [14] , [15] . It has also been proposed that zebrafish muscle cell elongation may be similar to notochord/neural plate cell intercalation [24] , represented by the protrusion scenario . Time-lapse analysis indicates that elongating cells extend local protrusions along their long axis ( Figure 8C , D ) . Protrusions are extended in the direction of elongation and between other cells . Protrusions then thicken , resulting in elongation of the cell . Repetition of protrusion extension/thickening results in an elongated muscle cell . Mathematical modeling of expected changes in area and perimeter supports the protrusion model of morphogenesis . Thus , we show that a novel two-step mechanism underlies elongation of the first fast muscle fibers in a vertebrate model system , the zebrafish . Muscle development is perhaps best understood in Drosophila , where muscle morphogenesis is accomplished via fusion of founder cells ( FCs ) with fusion competent myoblasts ( FCMs ) [13] . Recent 3-D imaging has demonstrated that there are two phases of fusion and suggests that the spatial relationship of FCs and FCMs influences the frequency of fusion events [12] . Exciting recent studies using zebrafish suggest that molecular events underlying muscle cell fusion in vertebrates may be at least partially conserved [14] , [15] , [40] . In the future it will be important to understand the cellular basis of fusion as well . In this regard , we show that elongating/recently elongated muscle cells possess complex 3-D shapes . Thus , a comprehensive analysis of cell behaviors underlying muscle cell fusion during zebrafish development will require development of multiple markers that label entire muscle cells such that fusion can unambiguously be analyzed . Genetic mosaic approaches such as those used previously [41] will facilitate analysis of both the timing of fusion as well as identifying what cells fuse . We show that lamβ1 and γ1 are required for efficient fast muscle cell elongation and proper organization . Application of the 2D WTMM method indicates that even in early stages of muscle development where organizational differences are not visually obvious , anisotropic signatures reveal unequivocally the morphological discrepancies between laminin-deficient and control embryos . This emphasizes the strength of the 2D WTMM method . This novel use of the 2D WTMM method will give researchers an invaluable tool to rigorously and quantitatively distinguish subtle differences in cellular morphology and organization . We do not know why fast muscle cell elongation is delayed in lamβ1 and γ1-deficient embryos . Elongation may be delayed because fast cells are less organized than in controls . It is also possible that fast cells in lamβ1 and γ1 mutant/morphant embryos do not elongate efficiently because slow muscle cells do not migrate efficiently . Although WT slow fibers can rescue elongation in mutant embryos that do not have slow muscle fibers [28] , it is unknown if disrupted slow muscle migration and/or morphology may delay fast muscle cell elongation . A third model is that adhesion to laminin may play a role in generation of traction forces that allow muscle cells to elongate . Muscle cells extend protrusions as they elongate and these protrusions likely attach to other cells or the ECM . Attachment would provide a mechanism for cells to stabilize an extended protrusion and continue elongation . Interestingly , adhesion to laminin via the Integrin α7β1 receptor promotes migration of C2C12 and MM14 cells in culture [42] . Elongating fast muscle cells in zebrafish do not migrate per se , but future studies will address whether adhesion to laminin during fast muscle cell elongation in zebrafish promotes efficient protrusion extension and thickening . These studies would be facilitated by identification of the relevant laminin receptor ( there are multiple laminin receptors ) such that genetic mosaic analysis could readily be used . Fast muscle cells do belatedly elongate in the absence of laminin . It is possible , even likely , that elongating muscle cells may utilize different modes of adhesion to the substrate and/or other cells . Thus , if one mode of adhesion is disrupted , muscle cell elongation would be delayed , but not entirely inhibited . Our results indicating that muscle cell elongation is delayed , rather than inhibited , are similar to the finding that myofiber formation is delayed , but recovers in mouse knockouts of the cell-cell adhesion protein CDO [43] . Taken together , these results suggest that muscle cells elongate by extension of protrusions that adhere both to other cells and the ECM . If one mode of adhesion is disrupted , cells are delayed in their elongation , but utilize the alternative mode of adhesion to eventually elongate ( Figure 8C , D ) . One fundamental process during embryonic development is boundary formation . Some of the first work describing boundary formation was done by Jacobson and colleagues [44]–[46] , where they showed that cells that reach the notoplate/neural plate boundary remain on the boundary permanently in both axolotl and newt embryos . This phenomenon was referred to as trapping . Keller and colleagues have since expanded upon this model and termed it boundary capture [47] . Recent work demonstrated that laminin plays a critical role in boundary capture during notochord morphogenesis in the ascidian Ciona savignyi [48] . We have previously demonstrated that the MTJ captures elongating muscle fibers , but it was not known what ECM components were relevant [24] . It was also not known if the cessation of muscle fiber elongation is cell autonomous or mediated by community effects . Here we show that laminin is one component of the MTJ that stops elongating fibers . This result , combined with the work of Veeman et al . , suggests that roles for laminin in boundary capture may be conserved , at least within chordates . We also show that wild-type cells in lamγ1 morphant embryos have a reduced ability to cross the MTJ . The fact that wild-type cells are less able to cross the MTJ , but do not rescue their lamγ1-deficient neighbors , suggests that boundary capture is a cell autonomous process . These data also suggest that MTJ breakdown in lamβ1and γ1-deficient embryos is a local event caused by the failure of elongating muscle fibers to stop when they reach the MTJ . We do not currently know why 75% of elongating muscle cells in lamβ1 and γ1-deficient embryos do stop elongating , but 25% do not . We hypothesize that the MTJ boundary is not homogenous . In this scenario , the absence of laminin would leave “holes” in the MTJ and muscle cells would elongate through these holes ( Figure 8B , B1 ) . Future experiments will be directed towards identifying additional molecular cues involved in boundary capture .
Zebrafish embryos were obtained from natural spawnings of adult fish kept at 28 . 5°C on a 16 h light/8 h dark cycle and were staged according to [49] . F59 was utilized to visualize slow fibers as previously described [25] , [50] . Alexa Fluor 488 and 546 phalloidin and Sytox green were obtained from Molecular Probes . We used the H2A∶GFP transgenic line of zebrafish to visualize nuclei [51] . A “scatter” label of cells filled with fluoro-ruby dextrans ( Molecular Probes ) was obtained by microinjecting embryos at the 512–1000 cell stage with dextrans into the yolk cell close to the margin . Antibodies used were: mouse monoclonal anti-myosin ( F59 ) ( Devoto , et al . 1996 , generous gift of Frank Stockdale ) 1∶10 , mouse monoclonal anti-β-catenin ( Sigma ) 1∶500 and Alexa-Fluor 488 , 546 and 633 conjugated goat anti-mouse and goat anti-rabbit secondary antibodies ( Invitrogen ) 1∶200 . Images were acquired using a Leica SP2 confocal microscope and a Zeiss ApoTome running on a Zeiss Axio Imager Z1 . All mathematical analyses were done on images acquired on the Apotome using a 20× lens , NA 0 . 8 , yielding a resolution of 1 . 5 pixels / µm . Images were linearly processed in Adobe Photoshop and collated in Adobe Illustrator . Morpholino-modified antisense oligonucleotides ( MOs ) were synthesized by Gene-Tools , LCC . The morpholinos used were previously described and recapitulate the mutant phenotypes [31] . Embryos were vitally stained and imaged with the fluorescent , lipophilic dye BODIPY-Ceramide ( Molecular Probes , Eugene , OR ) using the procedures outlined by [52] , [53] . Time-lapse recordings were made using a scanning laser confocal microscope ( Leica SP2 , Heidelburg , Germany ) . Time-lapse analysis with transplanted dextran-filled cells was performed utilizing the Zeiss ApoTome . To measure properties of dextran-filled cells , the z-series of the cell was projected such as to visualize the cell three-dimensionally . Cells were then segmented with ImageJ and the perimeter , area and major axis were measured . The major axis as determined by ImageJ is the longest length of the best fitting ellipse . The filament index was also calculated [54]:where P , D and A are the perimeter , diameter and area respectively . For this study , the diameter was taken to be equal to the major axis . Note that a circle has a filament index F = 1 and an object having a value of F larger than 1 quantifies its departure from a circular shape . A two sample T-test was performed using SYSTAT . * denotes p<0 . 05 and ** denotes p<0 . 01 . In order to quantitatively characterize the morphology of muscle fiber growth , two geometrical models were developed . For simplicity , both the elliptical and protrusion models start with a unit circle ( with radius = 1 ) . For both models , when the cell is growing we assume that it does so only in the major direction and that the semi-minor axis stays constant , and for mathematical simplicity ( and without any loss of generality ) , is equal to 1 . Therefore , we haveThe growth ratios for the area and perimeter of both models can be defined analytically . For the elliptical model , the growth ratio for the area at time t , Âelliptical ( t ) , is equal to the major axis at time t , which grows continuously:The perimeter growth ratio at time t , Pˆelliptical ( t ) , is given byFor the protrusion model , the cell grows in a two-step manner , expanding a thin protrusion of relatively small area ΔA and perimeter ΔP at time t , and then filling the area until the cell becomes an ellipse at time t+1 . Therefore , the growth ratio for the area of the protrusion model will depend on whether it is growing a protrusion or filling that protrusion . For simplicity , we assume that the cell is growing a protrusion if t is even and it is filling the area opened by protrusion when t is odd:Similarly for the perimeter of the protrusion model cell:Since ΔA is relatively small with respect to the area of the whole cell , the extension of a protrusion will not significantly increase the area of the cell . Conversely , since ΔP is relatively large with respect to the perimeter of the whole cell , the perimeter of the cell will significantly increase . The evolution of the area and perimeter as a function of time for both models is shown in Figure 3A , A1 . In order to obtain quantitative information from the angle pdfs Pa ( A ) , they are compared to a theoretical flat distribution representing an ideal isotropic signature ( see Figure 5F ) . The anisotropy factor , Fa , defined for each value of the scale parameter a , is given by the area between the curve corresponding to the observed pdfs and a flat distribution: Therefore Fa has been defined in such a way that a theoretically isotropic surface will have a value of Fa = 0 , while any value greater than 0 quantifies a departure from isotropy . Following the standard procedures presented in [34] , [35] , fractional Brownian motion ( fBm ) isotropic surfaces were generated . Two-dimensional fBm's are processes with stationary zero-mean Gaussian increments that are statistically invariant under isotropic dilations . They are therefore expected to reproduce quite faithfully the isotropic scaling invariance properties . WT embryos were injected with 10 , 000 MW dextrans ( Molecular Probes ) . Cells were removed at the sphere stage and placed into hosts that had been injected with lamγ1 MOs . Hosts were grown up until the appropriate stage , stained with phalloidin and the number of transplanted control cells that crossed MTJ boundaries was compared with the number of lamγ1 morphant cells that crossed MTJ boundaries . | Despite the importance of muscle fiber development and tendon attachment , this process is incompletely understood in vertebrates . One critical step is muscle fiber elongation; muscle precursor cells are short and subsequent elongation/fusion generates long , multinucleate muscle fibers . Using a vertebrate model organism , the zebrafish , we find that single round myoblasts elongate to span the entire width of the myotome prior to fusion . Using rigorous and objective mathematical characterization techniques , we can further divide muscle development into three stages: short precursor cells , intercalation/elongation , and boundary capture/myotube formation . The second phase , elongation , occurs via a two-step mechanism of protrusion extension and filling . Myotube formation involves boundary capture , where the ends of muscle fibers anchor themselves to the myotome boundary and stop elongating . We show that the protein laminin is required for boundary capture , normal fiber length , and proper fiber orientation . Genetic mosaic experiments in laminin-deficient embryos reveal that boundary capture is a cell autonomous phenomenon . Wild-type ( normal ) cells capture the boundary appropriately and stop elongating in laminin-deficient embryos . Although adhesion to laminin has been implicated in muscular dystrophies where the attachment between muscle cells and tendons fails , no early developmental requirements for laminin in fast muscle morphogenesis have been shown until now . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"developmental",
"biology/embryology",
"developmental",
"biology/morphogenesis",
"and",
"cell",
"biology",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"cell",
"biology/developmental",
"molecular",
"mechanisms",
"genetics",
"and",
"genomics/gene",
"function",
"m... | 2008 | Time-Lapse Analysis and Mathematical Characterization Elucidate Novel Mechanisms Underlying Muscle Morphogenesis |
Prevention is most effective in reducing dengue infection risk , especially in endemic countries like Pakistan . Evaluation of public awareness and health beliefs regarding dengue fever ( DF ) is important for devising disease control strategies . This study assessed dengue knowledge , health beliefs , and preventive practices against DF in different socioeconomic groups of Karachi , Pakistan . In this community-based cross-sectional study , 6 randomly selected towns were visited , 2 persons ( man and woman ) per household were interviewed using a structured questionnaire , and household practices were observed . Information regarding DF was shared through a printed pamphlet . Multivariate logistic regression analysis of variables associated with dengue knowledge and practices was conducted . We interviewed 608 Karachi residents ( mean age: 33 . 2 ± 13 . 35 years ) ; 7 . 7% , 71 . 9% , and 20 . 4% had a high , middle , and low socioeconomic status , respectively . The mean knowledge score was 6 . 4 ± 2 . 10 out of 14 . The mean preventive practices score was 9 ± 1 . 8 out of 17 . Predictors of dengue knowledge were perceived threat ( odds ratio [OR] = 1 . 802; 95% confidence interval [CI] = 1 . 19–2 . 71; p = 0 . 005 ) , self-efficacy ( OR = 2 . 910; 95% CI = 1 . 77–4 . 76; p = 0 . 000 ) , and television as an information source ( OR = 3 . 202; 95% CI = 1 . 97–5 . 17; p = 0 . 000 ) . Predictors of dengue preventive practices were perceived threat ( OR = 1 . 502; 95% CI = 1 . 02–2 . 19; p = 0 . 036 ) , self-efficacy ( OR = 1 . 982; 95% CI = 1 . 34–2 . 91; p = 0 . 000 ) , and dengue knowledge ( OR = 1 . 581; 95% CI = 1 . 05–2 . 37; p = 0 . 028 ) . Public knowledge about DF is low in Karachi . Knowledge , threat perception , and self-efficacy are significant predictors of adequate dengue preventive practices . Prevention and control strategies should focus on raising awareness about dengue contraction risk and severity through television . Health messages should be designed to increase individual self-efficacy .
Dengue fever is a massive health threat throughout the world , with the global estimate of the dengue-affected population reaching almost 400 million [1] . Factors augmenting dengue spread are uncontrolled urbanization , population growth , and lack of preventive measures in endemic areas [2] . A local study reported an incidence rate of 570/100 , 000 per year in the 10 to 15 years age group [3] . In 2011 , 22 , 562 dengue cases were confirmed , with 363 deaths recorded in the country . In the Sindh province alone , 952 cases were reported , with 18 deaths , of which 755 cases , including 15 deaths , were from the Karachi metropolis alone [4] . In 2015 , 3 , 212 cases were detected in Karachi , with an incidence rate of 35 . 6 per 100 , 000 in the 9-million population . Many additional cases are underreported or missed due to mild symptoms and inappropriate surveillance for dengue [5] . Dengue fever is a severe influenza-like infection that affects all age groups and rarely causes death , but in developing countries like Pakistan , dengue has the potential to cause high mortality because of an improper water system and sanitation , a large number of refugees , uncontrolled urbanization , improper urban infrastructure , frequent natural disasters , and a lack of resources [6] . The World Health Organization has been working with the Ministry of Health of Pakistan to cope with the situation . This includes epidemiological work on the dengue vector , the Aedes mosquito , in Karachi in 2005 , followed by the design of a Pakistan-specific control intervention plan . In 2008 , the plan was merged with the malaria control program to provide effective , long-term control and prevention of dengue . Despite these efforts , the incidence of dengue is still rising in many parts of the country [7] . Dengue prevention and control can be achieved by adopting proper preventive measures such as the use of mosquito nets , repellent spray , and mosquito coils; removal of stagnant water; and avoidance of trash accumulation in and outside houses to prevent mosquito breeding . In this regard , community awareness about dengue , its perceived severity and susceptibility , and practices and beliefs have a great impact on the prevention and control of dengue in the community [8] . There are factors like socioeconomic status , gender , literacy , and household income that influence the awareness/knowledge , beliefs , and preventive health behavior of an individual [9] . It is important to identify the influence of these factors according to the target population to form an effective plan for the control and prevention of any disease . Dengue knowledge , prevention practices , and the associated demographic factors in relation to Health Belief Model ( HBM ) constructs have never been explored in the population of the Karachi metropolis , Pakistan . In this study , using the HBM constructs , we attempted to find out individuals’ perceived dengue threat and their household practices for the prevention and control of dengue . The HBM constructs can be used to predict why people take action to control or prevent a particular illness or disease . These constructs are perceived threat of a particular condition , perceived benefits and barriers , perceived self-efficacy ( ability to avoid dengue through preventive practices ) , and cues to action ( measures that may increase awareness and readiness in executing preventive practices ) [10] . These factors could guide the design of dengue-related targeted interventions and the development of an effective educational/awareness program for the targeted population .
This was a cross-sectional community-based study . Out of the 18 towns of Karachi , 6 towns ( Gulberg Town , Liaquatabad Town , Saddar Town , Jamshed Town , Bin Qasim Town , and Kiamari Town ) were selected for data collection , using the lottery method . Each town is divided into a different number of areas . Through lottery , one randomly selected area was chosen from each town: Aisha Manzil ( Gulberg Town ) , Liaquatabad ( Liaquatabad Town ) , Clifton ( Saddar Town ) , Pakistan Employees Co-operative Housing Society ( Jamshed Town ) , Cattle colony ( Bin Qasim Town ) , and Maripur ( Kiamari Town ) . For each district , a list of mohala/sectors was acquired from the respective Union Council offices . Five sectors were chosen through lottery , and 20 houses from each selected sector were selected systematically ( every third house ) for data collection . Visits for data collection from systematically selected houses were done spontaneously , with respondents not expecting the visit . The data collection team comprised 4 men and 2 women . This study was approved by the institutional Ethics Review Committee of Jinnah Postgraduate Medical Centre , Karachi . The selection of participants was based on 1 ) participation in home chores and responsibility for making decisions concerning the health of the family , 2 ) age of 18 years or above , 3 ) provision of written informed consent for participation in the study , and 4 ) residence in the respective town for more than 5 years . According to the estimated sample size , a total of 608 respondents were recruited for this study . After obtaining verbal and written informed consent , 1 man and 1 woman from each house were interviewed using a structured pretested questionnaire . Household practices and cleanliness of the home were also observed and recorded . There were a total of 35 questions . The questionnaire was based on a literature review . It was validated by 6 experts who judged every item on the basis of its relevance and clarity . Each item was graded as “not relevant , ” “somewhat relevant , ” “quite relevant , ” or “highly relevant . ” The content validity index for the 35 items was calculated to be 0 . 89 . The questionnaire was pretested in the population . The questionnaire explored sociodemographic information , knowledge about dengue , experience of dengue , perceived threat of dengue , participant opinion on or understanding of the ability to effectively take measures to prevent dengue , i . e . , self-efficacy , and observation of practices related to dengue prevention and control . The following sociodemographic information was obtained from the participants: 1 ) age , 2 ) gender , 3 ) education , 4 ) socioeconomic status , 5 ) approximate household income , and 6 ) number of dependents . The socioeconomic status was determined on the basis of household income per person . A per person monthly income of <3 , 000 , 3 , 000–5 , 000 , and >5000 rupees ( PKRs ) was considered to indicate a low , middle , and high socioeconomic status , respectively . Knowledge about dengue was evaluated with questions on 1 ) dengue vector , 2 ) transmission of dengue , 3 ) dengue signs and symptoms , 4 ) places where dengue mosquitos can breed and measures to avoid breeding , and 5 ) treatment and preventive measures to avoid mosquito bites . All correct answers were scored as 1 , and incorrect answers and responses indicating uncertainty were scored as 0 . If a question had more than one correct answer , each of them was scored as 1 . The total score for the questions regarding knowledge about dengue was 14 . Knowledge about dengue was considered poor if the respondent scored ≤7 . A knowledge score between 8 and 14 was considered good . Participants’ understanding of/feeling of vulnerability to contracting dengue fever , called perceived susceptibility , and its seriousness , including health consequences , i . e . , perceived severity , were evaluated together as perceived threat . This was scored from 0 to 9 , with scores of 0 to 5 considered low and scores of 6 to 9 considered high for the purpose of the present analysis . In the present study , self-efficacy was measured on the basis of 3 activity questions . These questions were designed to assess the level of self-efficacy , i . e . , the number of activities an individual thought could be performed to avoid dengue infection . Participants who responded “Yes” for all activities were considered as confident to take measures to prevent dengue infection . Practices regarding dengue prevention and control were assessed by asking questions as well as by observation of household practices . The questions regarding practices were about the use of mosquito repellent , changing of water in flowerpots , the use of mosquito nets , and the removal of stagnant water from the roof as well as from any inside areas of the house . Observations were made regarding the cleanliness of the home; whether water containers were covered; and the presence of stagnant water near plants , in used bottle piles , in pets’ water containers , as well as in bathrooms . Each positive practice was scored as 1 , with a maximum possible total practice score of 17 . A score ≤9 was considered to indicate inadequate practices , and a score from 10 to 17 was considered to indicate adequate practices . After the interview and data collection , a dengue information pamphlet written in the native language was distributed among the respondents . In addition , the respondents were educated about dengue , preventive practices , and the individual’s role in combatting this infection through verbal communication . Data analysis was done on Microsoft Office Excel , using Statistical Package for Social Science , version 21 . The chi-square test was done to identify significant associations of dependent variables ( dengue knowledge and preventive practices ) with gender , age , literacy , socioeconomic status , monthly income , perceived threat , and self-efficacy . For all statistical analyses , a p-value less than 0 . 05 was considered to indicate statistical significance . The association between sources of information and dengue knowledge , preventive practices , perceived threat of dengue , and self-efficacy was also analyzed . Logistic regression analysis was done to define the partial contribution of each independent variable to dependent variables like preventive practices score and knowledge score . Variables with a p-value less than 0 . 05 were included in logistic regression analysis . Variation in the dependent variable was assessed using Cox-Snell R-square . Nagelkerke’s R-square was also used to measure the relationship between predictors and predictions . Model chi-square and Hosmer-Lemeshow tests were also used to check the goodness of model fit .
In total , 311 households from all the 6 selected towns were approached . Seven ( 2 . 3% ) households refused to participate in the study . Responses of a total of 608 participants were recorded . The mean age of study participants was 33 . 2 ( ±13 . 35 ) years; 302 ( 49 . 6% ) were male and 306 ( 50 . 3% ) were female . Out of the 302 male respondents , 158 ( 52 . 3% ) were reported to be the head of the household . A total of 68 ( 22 . 2% ) female respondents out of 306 were elderly women; the remaining women were involved in household chores . There were 231 ( 76% ) male and 228 ( 74% ) female literate respondents . Overall , out of these 608 participants , 47 ( 7 . 7% ) , 437 ( 71 . 9% ) , and 124 ( 20 . 4% ) belonged to the high , middle , and low socioeconomic status groups , respectively . Among these respondents , 149 ( 24 . 5% ) were illiterate or completed less than 10 grades , whereas 459 ( 75 . 5% ) completed 10 or more grades and were considered literate . A total of 417 ( 68 . 6% ) reported a household monthly income of PKRs 20 , 000 or less , and 191 ( 31 . 4% ) reported a household monthly income of more than PKRs 20 , 000 . Analysis of associations of the sources of dengue information with dengue knowledge , preventive practices , perceived threat of dengue , and self-efficacy showed a significant association for television , newspapers , and the government campaign performed to promote dengue knowledge in the community ( Table 1 ) . The government campaign was the “Dengue Prevention and Control Program , ” which has been launched to provide surveillance data , spray insecticides , and improve public knowledge to implement adequate dengue preventive practices . A total of 522 ( 85 . 9% ) respondents claimed that they knew about dengue infection . A total of 517 ( 85% ) knew that it is spread by mosquitos . The main symptoms of dengue were reported as headache by 174 ( 28 . 6% ) , body pain by 284 ( 46 . 7% ) , and eye pain by 80 ( 13 . 2% ) ; 440 ( 72 . 4% ) indicated that dengue infection can occur in all age groups . Only 198 ( 32 . 6% ) were aware that dengue fever mostly occurs in the monsoon ( rainy ) season , and 419 ( 68 . 9% ) reported that dengue mosquitos breed in fresh water . One hundred eighty-six ( 30 . 6% ) , 173 ( 28 . 5% ) , and 240 ( 39 . 5% ) respondents agreed that used tires , used/spare boxes , and water present in flower pots/vases are sites for mosquito breeding . Almost all the respondents , 586 ( 96 . 4% ) , agreed that water at home should be stored in a covered utensil/tank , and 344 ( 56 . 6% ) reported that uncovered water utensils should be cleaned on a daily basis . A total of 353 ( 58 . 1% ) stated that this is a contagious infection that can be transmitted from one person to another . The mean score for knowledge about dengue fever was 6 . 4 ± 2 . 10 out of 14 . As shown in Table 2 , there were significant associations between dengue knowledge and age , literacy , history of having dengue fever in the last 2 years , perceived threat , and self-efficacy . We found it interesting that those who had suffered from dengue fever in the past 2 years had a significantly lower knowledge score . A further , separate analysis of a past dengue history with perceived threat of dengue , self-efficacy , and practices revealed that perceived threat and self-efficacy were significantly associated with a history of dengue fever ( p < 0 . 01 ) ( Fig 1 ) . When we compared the respondents’ willingness to participate in the government campaign with health beliefs , knowledge , and practices , a significant association ( p < 0 . 05 ) was found with these variables ( Fig 2 ) . Before regression analysis , its assumptions were verified . The Durbin-Watson test statistic was used for verification of the independence of error assumption . The value of the Durbin-Watson test statistic was within the acceptable range ( 1 . 434 and 1 . 511 ) . Thus , we failed to reject the null hypothesis that there was no correlation among residuals , i . e . , they were independent . In multiple logistic regression analysis , the model chi-square was statistically significant ( chi-square = 104 . 77 , p < 0 . 000 with df = 9 ) . The Wald criterion demonstrated that only perceived threat , agreement on self-efficacy , and television as a source of dengue information made a significant contribution to the prediction of high dengue knowledge ( Table 3 ) . Self-efficacy ( odds ratio [OR] = 2 . 910; 95% confidence interval [CI] = 1 . 77–4 . 76; p = 0 . 000 ) and television ( OR = 3 . 202; 95% CI = 1 . 97–5 . 17; p = 0 . 000 ) as a source of information were associated with higher ( 3-fold ) odds of adequate dengue knowledge . Higher perceived threat scores ( OR = 1 . 802; 95% CI = 1 . 19–2 . 71; p = 0 . 005 ) were associated with higher ( 2-fold ) odds of adequate dengue knowledge . Age , literacy , history of dengue fever in the last 2 years , and sources of information , i . e . , newspapers and the government campaign , were not significant predictors of dengue knowledge . Regarding dengue practices , 292 ( 48% ) , 230 ( 37 . 8% ) , 138 ( 22 . 7% ) , and 109 ( 17 . 9% ) reported that mosquito eradication , use of mosquito nets , body covering with cloths , and application of mosquito repellent lotions were effective in avoiding mosquito bites . Most of the respondents , 543 ( 89 . 3% ) , covered water storage utensils . Further , 399 ( 65 . 6% ) respondents stated that they changed flower pots/vases daily , only 125 ( 20 . 6% ) used mosquito nets , 320 ( 52 . 6% ) used coils , 148 ( 24 . 3% ) used mosquito repellent , 152 ( 25% ) used mosquito sprays at home , and a few of them , 44 ( 7 . 2% ) , also used smoke to avoid mosquito bites . Household observation showed that the houses of 572 ( 94 . 1% ) respondents were clean . Flower pots and plants were clean in the houses of 489 ( 80 . 4% ) respondents , uncovered water pots were seen in the houses of 176 ( 28 . 9% ) respondents , and 189 ( 31 . 1% ) respondents had empty pots and boxes in which water could be stored and could serve as a place for mosquito breeding . Three hundred twenty ( 52 . 6% ) respondents had clean uncovered pet water containers . The mean score of preventive practices was 9 ± 1 . 8 out of 17 . Dengue preventive practices were significantly associated with age , literacy , perceived threat , self-efficacy , and dengue knowledge ( Table 4 ) . Regarding sources of dengue information , we found that television , newspapers , family , as well as the government campaign were significantly associated with dengue preventive practices ( Table 1 ) . In logistic regression analysis , the model chi-square was statistically significant ( chi square = 61 . 98 , p < 0 . 000 with df = 9 ) , indicating an overall significant model . The Wald criterion indicated that only perceived threat , self-efficacy , and dengue knowledge significantly contributed as true predictors of dengue preventive practices ( Table 5 ) . It was observed that higher scores for perceived threat ( OR = 1 . 502; 95% CI = 1 . 02–2 . 19; p = 0 . 036 ) were associated with higher ( 2-fold ) odds of dengue preventive practices . Similarly , participants with higher knowledge ( OR = 1 . 581; 95% CI = 1 . 05–2 . 37; p = 0 . 028 ) and self-efficacy ( OR = 1 . 982; 95% CI = 1 . 34–2 . 91; p = 0 . 000 ) scores had higher ( 2-fold ) odds of implementing dengue preventive practices . Other candidate factors like age , literacy , and sources of information were not significant predictors for adequate dengue preventive practices .
In this study , most of the respondents , 522 ( 85 . 9% ) , believed that they have knowledge about dengue . Further evaluation revealed that in Karachi , an endemic area for dengue , the mean score of dengue knowledge is very low , i . e . , 6 . 4 ± 2 . 10 out of 14 . The rate of adequate dengue knowledge within the community was only 32 . 4% . This is comparable with the results of a local study published in 2010 [11] but inconsistent with the study conducted in “Wah Cantonment” Punjab province that showed a 54 . 3% rate of adequate knowledge [12] . A community survey from a neighboring country Nepal reported a 12% rate of adequate knowledge [13]; our findings are quite fair in comparison to this figure but not satisfactory . Our findings showed that , except for age and literacy , participants’ demographic characteristics are not significantly associated with dengue knowledge ( Table 2 ) . In past studies conducted in Karachi , literacy as well as socioeconomic status was significantly associated with adequate dengue knowledge [11 , 14] . We have not found any significant association between socioeconomic status of the respondent and dengue knowledge . This study showed that the respondent’s age has a significant association with dengue knowledge , with 37 . 7% of the respondents in the older age group , i . e . , 36–55 years , having adequate knowledge as compared to 29 . 2% in the younger age group , i . e . , 15–35 years . Similar results have been obtained in the study conducted in Wah Cantonment [12] . The lack of knowledge in the younger age group indicates that the media and sources of information about dengue might not be adequately reaching this age group compared to older individuals , who willingly show interest and seek information . We found that those who had had dengue fever in the past 2 years had a significantly lower level of dengue knowledge , which contradicts a previous study that showed a significant impact of past dengue exposure on adequate dengue knowledge [9] . From this finding , we may surmise that , in our setup , health care providers do not adequately provide useful information about dengue prevention to their patients . Interestingly , our analysis of a history of dengue in the previous 2 years with the HBM constructs of perceived threat and self-efficacy revealed that it was negatively associated with these constructs . We may surmise that such individuals might be at a higher risk of contracting dengue due to their low perceived threat and self-efficacy . Preventive practices for dengue were also associated with a history of dengue fever in the past 2 years , with the data indicating that adequate dengue preventive practices are very high in those who do not have a past history ( Fig 1 ) . On regression analysis , none of these demographic factors significantly contributed as an independent predictor of adequate dengue knowledge in our population . We also tried to determine positive associations of willingness to support the government campaign for dengue with public knowledge about dengue , practices , and health beliefs , and a significant association was found for all these variables . Almost all respondents who had high scores for these variables were willing to participate in and support the government campaign for dengue prevention in their area ( Fig 2 ) . No association was found between respondents’ willingness and their age , gender , or socioeconomic status . Electronic and print media are considered to play an important role in the dissemination of information to the public . Regression analysis showed that radio , newspapers , and others sources of information about dengue are not significant predictors compared to television . Regression analysis suggested that television as an information source was associated with higher ( 3-fold ) odds of knowledge about dengue compared to other sources of information . These findings support other local and international studies that identified television as a useful source of information [14 , 15] . The reason for information delivery via television being a significant predictor of adequate dengue knowledge could be that television is the most popular form of media that appeals to every socioeconomic class , including the literate and the illiterate , and every age group in our part of the world . Our finding suggests that television as a source of information can be used as an effective means to promote dengue awareness among the masses . Regarding the HBM constructs , we found that higher perceived self-efficacy and threat scores are associated with higher ( 2–3-fold ) odds of adequate dengue knowledge . These findings are in concordance with a recent study carried out in Malaysia [16] . According to the HBM , perceived threat encompasses 2 main HBM constructs , perceived susceptibility and perceived severity . From these findings , we may infer that a person who feels susceptible to dengue and believes that contracting dengue may lead to severe health consequences will consequently be active in seeking more information about dengue in order to prevent this disease . In addition , this feeling of “self-efficacy , ” i . e . , the realization that “on adopting precautionary measures this disease can be prevented , ” supports information-seeking behavior , eventually leading to adequate dengue knowledge . It is interesting that perceived threat was significantly associated ( p = 0 . 000–0 . 007 ) with all sources of information used to increase dengue knowledge ( Table 1 ) . We could say that this important HBM construct , perceived threat , can easily be comprehended by our population , regardless of the information source used . The situation is different for self-efficacy , as only television and newspapers showed a significant association with this construct ( Table 1 ) . Our findings suggest that the information about dengue disseminated in the community should focus on increasing public perceived threat and perceived susceptibility to acquiring this disease , as well as on the understanding of self-efficacy and assurance that this infection can be prevented by effective preventive practices . With reference to the HBM , framing health messages for dengue prevention according to HBM constructs could result in an effective dengue control program in Karachi [17] . Although the HBM is a conceptual guiding framework for health behavior intervention , it has some limitations . For example , the HBM does not account for environmental factors that may prevent an individual from practicing the desired behaviors . For instance , an individual’s determination to adopt dengue preventive practices may be limited by poor infrastructure , bad sanitation , and bad water supply . In addition , the HBM does not consider the emotional component of behavior . Furthermore , “cues to action” cannot be assessed precisely by the HBM , as it is difficult for an individual to remember the cue that prompted the change of behavior . To the best of our knowledge , this is the first study from Karachi , Pakistan , that was carried out in the community and that recorded the actual practices of dengue prevention being carried out in households of this dengue-endemic metropolis . It was found that adequate dengue preventive practices were adopted in the households of 363 ( 59 . 7% ) respondents in the study population . This figure is quite impressive when compared with the 32 . 4% rate of adequate dengue knowledge in the population . Similar differences have been found in a national survey conducted in Malaysia [9] . These differences might be noted , as knowledge was measured using items that were beyond the scope of dengue preventive practices , i . e . , the Aedes mosquito , dengue symptoms , seasons , and medication . Moreover , some practices might be performed without knowledge about dengue just because of social acceptability or social norms , like covering utensils , keeping the house clean and clutter free , and removing empty boxes for the sake of cleanliness . Our results showed that age and literacy were significantly associated with dengue preventive practices . Adequate dengue prevention practices were used more often in the 36–55 years age group ( 151 [65 . 4%] ) than in the 15–35 years age group ( 212 [56 . 2%] ) . Similarly , literate respondents were better at adopting adequate dengue prevention practices ( 286 [62 . 3%] ) than illiterate respondents ( 77 [51 . 7%] ) . In a study conducted in Punjab , literacy had no association with actual prevention practices [12] . The inconsistency in the results might be due to the differences in population , as Karachi is a multilingual city where population characteristics vary more broadly than in other cities of Pakistan . Regression analysis showed that none of the demographic and other general features of respondents was found to be a significant predictor of adequate dengue preventive practices . The present study showed that perceived threat , self-efficacy , and knowledge about dengue are the only significant predictors of adequate dengue preventive practices ( Table 5 ) . Perceived threat is one of the important factors that increase readiness and motivation to take precautionary measures [18] . Our data suggest that individuals who have sufficient perception of susceptibility to dengue and feel the threat of acquiring the disease have higher odds of adopting preventive practices than others . This is in concordance with the study carried out in the Malaysian population [9] . Regression analysis also showed that those who have a perception of self-efficacy have higher odds of adopting adequate dengue practices than those who do not have confidence that they can prevent dengue through effective preventive measures . Self-efficacy is another HBM construct that , in addition to perceived threat and other constructs , encourages an individual to implement preventive practices [19] . It was found that dengue knowledge is also a significant independent predictor of dengue preventive practices . This is not consistent with the results of a study carried out in Lahore , another dengue-endemic city of Pakistan [20] , as well as with a study conducted in the Philippines [15] . Our findings are consistent with the outcome of studies carried out in Malaysia and Cuba [9 , 21] . We may infer from our findings that public health beliefs have a great impact on dengue preventive practices . Knowledge about dengue should be disseminated at a mass level so that preventive practices can be improved in the population . From Pakistan , this study is one of its kind in that it evaluated public knowledge about dengue and preventive practices by using the HBM . A limitation of the study is that the questionnaire did not include questions regarding perceived benefits and barriers related to dengue prevention . Thus , we were not able to analyze these important factors , which could also be useful to evaluate public perception of dengue prevention . We tried to minimize the biases related to self-reporting of dengue preventive practices through direct observation . From this study , we may conclude that knowledge about dengue is rather limited in the population of dengue-endemic Karachi , a metropolitan city of Pakistan . The present study showed that adequate dengue knowledge leads to adequate dengue preventive practices . Socioeconomic status and other demographics are not significant predictors of dengue knowledge and practices in the population . Older and literate individuals have more knowledge about dengue and adopt more preventive practices . Health beliefs are a significant predictor of both adequate dengue knowledge and adequate dengue preventive practices in our population . Hence , it is recommended that health messages and awareness campaigns should be formulated on the basis of these health belief constructs . There is a need for mass campaigning on television to disseminate dengue information emphasizing public susceptibility to dengue and the consequences of contracting this disease . The public message should be modified so that it increases individuals’ self-efficacy , such that they can avoid the contraction of dengue by adopting proper preventive practices . | Dengue is a massive health threat worldwide . Dengue prevention is the most effective way to reduce the risk of dengue infection , especially in endemic countries like Pakistan . Evaluation of public awareness and health beliefs regarding dengue fever plays an important role in developing strategies for disease control . The present study results highlight the public awareness level , health beliefs , and actual preventive practices regarding dengue fever . We evaluated dengue knowledge and preventive practices in relation with demographic characteristics and public health beliefs . Our findings are useful for developing health messages regarding dengue prevention and control . We found that public awareness has a great impact on adoption of dengue preventive practices . We also found that dengue information should be disseminated in the population in order to increase the perception of susceptibility to contracting this disease . It is also important that health messages include information that augments public confidence that implementation of proper preventive practices can help avoid this infection . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"demography",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"neuroscience",
"health",
"care",
"regression",
"analysis",
"age",
"groups",
"academic",
"skills",
"cognitive",
"psychology",
"mathematics",
... | 2016 | Use of the Health Belief Model for the Assessment of Public Knowledge and Household Preventive Practices in Karachi, Pakistan, a Dengue-Endemic City |
Understanding the factors that promote , disrupt , or shape the nature of cooperation is one of the main tasks of evolutionary biology . Here , we focus on attitudes and beliefs supportive of in-group favoritism and strict adherence to moral consensus , collectively known as ideological rigidity , that have been linked with both ends of the political spectrum . The presence among the political right and the left is likely to make ideological rigidity a major determinant of the political discourse with an important social function . To better understand this function , we equip the indirect reciprocity framework – widely used to explain evaluation-mediated social cooperation – with multiple stylized value systems , each corresponding to the different degree of ideological rigidity . By running game theoretical simulations , we observe the competitive evolution of these systems , map conditions that lead to more ideologically rigid societies , and identify potentially disastrous outcomes . In particular , we uncover that barriers to cooperation aid ideological rigidity . The society may even polarize to the extent where social parasites overrun the population and cause the complete collapse of the social structure . These results have implications for lawmakers globally , warning against restrictive or protectionist policies .
Factors affecting cooperation in a society , such as attitudes , beliefs , and resulting value systems , are a subject of major interest in evolutionary biology . Some examples of considerable importance are in-group favoritism , nationalism , ethnocentrism , intolerance for dissent , submission to strong leadership , and support for tight control all of which were originally linked with the political right [1] . However , evidence was presented to dispel such a link and argue that the same set of attitudes , by serving both the right and the left , is indicative of ideological rigidity rather than a position on the political spectrum [2] . The alleged presence on both ends of the spectrum is likely to make ideological rigidity a potent force in directing the political discourse and ultimately shaping societies . Here , we set to investigate the social function of ideological rigidity , starting from a motivational premise that indirect reciprocity – a cooperation maintaining mechanism based on the evaluation of the reputation [3] – provides a proper framework for our investigation . To establish this premise , we emphasize the dual nature of the aforementioned attitudes and beliefs . On the one hand , ideologically rigid believe in the supremacy of one's group or , at least , distrust anyone who is not a member of this group . Such a belief , broadly termed in-group favoritism , represents an attractive phenomenon for the studies on indirect reciprocity [4]–[6] . On the other hand , ideological rigidity is linked with attitudes that demand strict adherence to moral consensus . Corresponding ideas are again found in the indirect reciprocity framework , where social norms subjected to the evolutionary competition [7] , [8] handle dissent from moral consensus in different ways . Our aim is to unify these ideas by incorporating the dual nature of ideological rigidity into stylized value systems and then examine the consequent evolutionary dynamics . As the first step forward , we formalize the notion of ideological rigidity within the indirect reciprocity framework . In an indirect reciprocity game , members of a society , or players , encounter each other randomly , whereupon one player takes the role of a donor , while the other acts as a recipient . The donor can choose between two actions contingent on the recipient's reputation . By cooperating with the recipient , the donor incurs a cost , but the recipient benefits from a payoff for a net gain of for the society . By defecting , the donor avoids the cost , the recipient gains nothing , and the society is exactly where it was before the encounter . Every action is scrutinized by observers who assign the donor an appropriate reputation for the next round of the game . Maintaining a favorable reputation improves the prospects of receiving the payoff afterwards , thus justifying the willingness to incur the cost in the first place . The payoff is potentially received in the next round of the game from a third player – hence the name indirect reciprocity – who then serves as a donor , while the current donor takes the role of a recipient . The detailed rules governing which action should be taken and how the reputation should be assessed are called action-assessment strategies and represent a stylized version of the donor's value system . Action-assessment strategies are a part of the central process called the reputation dynamics ( see Methods ) . For now , it is critical that the reputation dynamics can incorporate several action-assessment strategies , allowing us to distinguish between player types and place them appropriately on a scale of ideological rigidity . The first key aspect of any action-assessment strategy , as the name suggests , is the action rule . We focus on the situation in which all players are discriminators , meaning that donors cooperate only with recipients who have a favorable reputation [9] . Because the action rule is the same for all players , making a distinction between player types requires other key aspects of action-assessment strategies to be more elaborate . One such aspect is the reputation assessment rule . Assessment rules are theoretical representations of social norms that govern the decision-making process of observers while assigning the reputation to donors for the next round of the game . We assume that information spreads from observers to other players rapidly ( e . g . through gossip ) . Two assessment rules are considered . The first of the two rules is called simple-standing or the Sugden rule [10]–[12] . It stipulates that a favorable reputation is assigned to a donor who cooperates with a recipient of favorable reputation or defects from a recipient of unfavorable reputation . An unfavorable reputation is assigned to a donor who defects from a recipient of favorable reputation . Importantly , a favorable reputation is assigned to a donor who cooperates with a recipient of unfavorable reputation , indicating that the Sugden rule liberally follows moral consensus . By contrast , the second of the two rules , called stern-judging or the Kandori rule [12] , [13] strictly enforces moral consensus . Cooperation with a recipient of unfavorable reputation leads to an unfavorable reputation assignment for the donor . For an easy comparison , both assessment rules are summarized in Table 1 . These concepts are defined in a strict mathematical manner in the section on the reputation dynamics ( see Methods ) . Before introducing another key aspect of an action-assessment strategy , we make the assumption that the society consists of two separate parts . Namely , an inner circle ( e . g . a nation state ) is embedded into a much larger outer world ( e . g . the international community ) , where the cooperation between the two parts of the society , though allowed , is made difficult ( e . g . by the national border control ) . Members of the inner circle ( i . e . insiders ) thus have a high probability , denoted , of encountering other insiders , but only a small probability , , of meeting a player from the outside world ( i . e . an outsider ) . For an insider , cooperation with the outside world also carries an additional cost , ( e . g . a tariff ) . Herein , we are primarily interested in the evolutionary dynamics ( see Methods ) of value systems inside the inner circle and the subsequent implications for ideological rigidity of insiders . The division of the society into two separate parts leads us naturally to another key aspect of an action-assessment strategy . Within the inner circle , because encounters with the outside world are rare , it is fairly reasonable to presume that distrust towards outsiders can take root among a fraction of the insiders . These insiders exhibit strong in-group favoritism in the sense that all cooperation with outsiders is suspended and no benefits from the outside world are accepted . The remaining insiders , by contrast , reject in-group favoritism , maintain cooperation with the outside world , and receive the accompanying benefits . The setting we describe here is not without a historical precedent . A resemblance can be found in pre-modern Japan [6] , where the two dominant value systems , one called bushido ( the way of warriors ) and the other called shonindo ( the way of merchants ) , held opposing positions on in-group favoritism . In Western culture , many parallels can be drawn by examining the differences between the Maghribi and the Genoese [14] . However , being primarily motivated by the bushido-shonindo dichotomy , we name the fraction of the insiders that embrace in-group favoritism “bushi” . The remaining insiders that reject in-group favoritism are named “shonin” . The two introduced aspects of action-assessment strategies ( Sugden vs . Kandori and shonin vs . bushi ) allow us to distinguish four types of insiders . Sugden-shonin ( hereafter Ss ) liberally follow moral consensus , reject in-group favoritism , and hence are considered ideologically non-rigid . A step up on the scale of ideological rigidity are Kandori-shonin ( Ks ) , who strictly enforce moral consensus , yet reject in-group favoritism . Sugden-bushi ( Sb ) , while liberal towards moral consensus , endorse in-group favoritism . We consider the stance of Sb players to be more ideologically rigid than the stance of Ks players because in-group favoritism as defined herein limits the scope of cooperation far more strongly than the strict enforcement of moral consensus . The most ideologically rigid are Kandori-bushi ( Kb ) , who enforce moral consensus and embrace in-group favoritism . Besides these four types of players we entertain the notion of social parasites in the form of unconditional defectors ( often denoted AllD in the literature , hereafter simply Ad ) . For the outside world , which is much bigger than the inner circle , interactions with insiders are inconsequential . Modeling the evolutionary dynamics ( see Methods ) of value systems in the outside world is possible using the same mathematical framework as for the inner circle , but with the probability of an outsider meeting another outsider set to unity . The inner circle is , therefore , a set of measure zero . Because we are interested in the evolutionary dynamics of value systems in the inner circle , the outside world is assumed to be in a stable equilibrium populated only by Ss or Ks players . Such a simple structure of the outside world can be justified by the fact that any other more complex structure would only diminish the benefits from cross-border encounters which is qualitatively captured by increasing the value of the parameter .
We explore the dual nature of ideological rigidity and its social function by means of indirect reciprocity games . Differences in adherence to moral consensus are reflected in the performance of the more liberal Sugden against the stricter Kandori rule ( S and K in shorthand notation , respectively ) . Similarly , opposite attitudes towards in-group favoritism are reflected in the performance of open-minded shonin against distrustful bushi players ( s and b , respectively ) . The focus is placed on the most illustrative cases , meaning a relatively closed inner circle in which the set of possible player types is either or . Though the model can handle any number of player types , having three types per simulation permits effective visualization and comparison of the results . We start with a technical description of two opposing situations , one where barriers to cooperation are low and the other where barriers are high ( Fig . 1 ) . To achieve this , we set both cost-benefit ratios , and , close to zero and subsequently increase either one towards unity . The results are then generalized by continuous mapping of the parameter space ( Fig . 2 ) and finally by inclusion of social parasites ( Ad ) into simulations ( Fig . 3 ) . Low barriers to cooperation favor ideologically non-rigid Ss strategy ( Figs . 1a , b ) . From ternary plots it is apparent that the vertices Ss and Ks share the property of being locally stable monomorphic attractors . By contrast , neither Kb nor Sb vertices have this property , but rather the whole segment connecting them is a locally stable dimorphic attractor ( the Kb-Sb attractor ) . Comparing the sizes of the corresponding domains of attraction reveals the evolutionarily advantageous action-assessment strategy . When cost-benefit ratios are close to zero ( Fig . 1a ) , the Ss attractor not only overshadows the Kb-Sb attractor in terms of the size of the domain of attraction ( 73% vs . 27% ) , but a rare occurrence of Ss players in an inner circle dominated by Sb players leads to a successful invasion . Under the same conditions ( Fig . 1b ) , the Ks attractor fares less well , commanding a smaller domain of attraction than the Kb-Sb attractor ( 46% vs . 54% ) and failing to successfully invade the inner circle dominated by any combination of Kb and Sb players . Increasing barriers to cooperation make the ideologically non-rigid Ss strategy evolutionarily disadvantageous ( Figs . 1c , d ) . For a society to maintain feasible cross-border interactions , ideological non-rigidity needs to be abandoned in favor of a more ideologically rigid Ks strategy . Namely , when the cost-benefit ratios is set close to unity , the domain of attraction of the Ss attractor ( Fig . 1c ) is greatly reduced in favor of the Kb-Sb attractor ( <1% vs . >99% ) despite Ss players still being able to invade an Sb-dominated inner circle . By contrast , the Ks attractor fares much better than originally ( Fig . 1d ) . Its domain of attraction is now larger than that of the alternative ( locally stable monomorphic ) Kb attractor ( 62% vs . 38% ) and a rare occurrence of Ks players in an Sb-dominated inner circle leads to a successful invasion . It is worth emphasizing that vulnerability to invasion by both Ss and Ks strategies makes the Sb strategy a weak candidate for the ideologically rigid . The increasing cost of cross-border interactions aids ideological rigidity ( Figs . 1e , f ) . The effect is twofold because the Sb strategy turns evolutionarily viable and the Ks strategy gains an ( albeit marginal ) evolutionary advantage over the Ss strategy . When the ratio is set close to unity and is kept near zero , the benefit of encountering outsiders is reduced and , therefore , Ss and Ks strategies are negatively impacted . Accordingly , the inner circle dominated by Sb players can no longer be invaded by either Ss or Ks players . The domains of attraction of Ss and Ks attractors become smaller than that of the Kb-Sb attractor ( 36% vs . 64% and 38% vs . 62% , respectively ) . A notable difference between Ss and Ks strategies , however , is that the negative impact of high is worse for the former ( Figs . 1a , e ) than the latter ( Figs . 1b , f ) strategy . When is close to unity the domain of attraction of the Ss attractor is reduced to half its original size ( from 73% to 36% ) , whereas the reduction for the Ks attractor is minimal ( from 46% to 38% ) . The three extreme cases examined heretofore are illustrative , but not exhaustive because any combination of cost-benefit ratios that satisfies is reasonable . Consequently , we map the parameter space in a continuous manner , focusing in particular on the relative performance of Ss and Ks strategies . We emphasize the relative performance because the evolutionary advantage of one strategy over the other changes with the location in the parameter space . By comparison , vulnerability to invasion suggests that for the ideologically more rigid the Sb strategy is a weak alternative . To summarize the relative performance of the two strategies over a wide range of cost-benefit ratios , we denote respectively by and areas of the domains of attraction corresponding to Ss and Ks attractors and introduce a performance indicator . Note that , where positive ( negative ) values indicate the evolutionary advantage of Ss ( Ks ) players . The simulation results ( Fig . 2 ) confirm the notion that Ss players have an evolutionary advantage when barriers to cooperation are low; that is , when the sum of the two cost-benefit ratios is close to zero . As barriers become higher , the Ks strategy turns out to be advantageous . Particularly detrimental for Ss players is the increase in the cost-benefit ratio because as it approaches unity the area tends to zero . Increasing affects both Ss and Ks players negatively , but the area is much more sensitive to the change in than . Pursuing a technical description of the simulation results so far sheds new light on the evolution of value systems as implied by the indirect reciprocity framework , but remains silent on the underlying mechanisms . We are thus required to make an extra effort to access these mechanisms and in return gain an intuitive grasp of the mathematical formalism being applied . Starting from a comparison of bushi and shonin players , the former are at a fundamental disadvantage because of refusing to interact with the outsiders . Such a situation is exemplified by the success of Ss over Sb players in Fig . 1a . Bushi , in fact , may not represent an evolutionarily viable alternative at all without sufficiently closed borders ( i . e . slightly below unity ) . If this criterion is met , the difference between Sugden and Kandori rules plays a major role . Players adopting the Sugden rule , due to their liberal stance , receive benefits and incur costs of cooperation more often than players adopting the Kandori rule – a clear advantage when is low . As approaches unity , the fortunes reverse . We can now understand why Sb players can resist Ks players in Fig . 1b , as well as the success of Kandori players in Figs . 1c and d . Protectionism ( i . e . slightly below unity ) naturally helps bushi agenda , which is best illustrated by how Ss players lose their advantage over Sb players as the cost of cross-border interactions increases ( Figs . 1a , e ) . The same effect is visible by comparing Figs . 1b and f , although it is much weaker because the Sugden rule to a certain extent shields Sb against Kb players even when the cost of cross-border cooperation is low . The described mechanisms can be used to explain all intermediate outcomes in Fig . 2 . Having determined that the Kb strategy shields the ideologically rigid from invasion , while the ideologically less rigid should choose between Ss and Ks strategies contingent on how high barriers to cooperation are set , we consider the effect of social parasites on the society . Indirect reciprocity games with two sets of possible player types , and , reveal important qualitative differences ( Fig . 3 ) . With moderately low barriers to cooperation , the conflict between Ss and Kb strategies in the presence of social parasites results in three distinct domains of attraction accompanied with three locally stable monomorphic attractors , as well as three dimorphic and one trimorphic equilibria ( Fig . 3a ) . Though the size of each domain of attraction is parameter-dependent , the remarkable outcome is that there are no openings for an invasion . By contrast , the conflict between Ks and Kb strategies in the presence of social parasites lacks a trimorphic equilibrium and leaves the dimorphic equilibrium of Ks and Kb players vulnerable to invasion in the case of a rare occurrence of Ad players ( Fig . 3b ) . Therefore , by making the ideologically more rigid Ks strategy evolutionarily advantageous over the Ss strategy , rising barriers to cooperation not only aid ideological rigidity , they even threaten the collapse of the social structure . Looking at the results in Fig . 3 , what we truly observe are the negative consequences of the maxim “the enemy of my enemy is my friend” as well as the way to avoid these consequences . When the inner circle is populated with an ideologically more rigid combination of Ks and Kb ( along with Ad ) players , even if they initially treat each other favorably , after a while a Ks player will cooperate with an outsider and be assigned an unfavorable reputation by the Kb observer . Such a player is bound to be denied cooperation from a Kb donor , resulting in an unfavorable reputation assignment for this donor from the Ks observer . A rift between Ks and Kb players forms . Ad players may eventually take the advantage of such a rift ( Fig . 3b ) because when they defect from a Ks recipient they receive benefits from a Kb donor and vice versa . Note that with stern-judging the rift only widens after every interaction of either a Ks or a Kb donor with an Ad recipient . The reason is that the Ad recipient is treated favorably by one side and unfavorably by the other , resulting with certainty in an unfavorable reputation assignment for the donor . Replacing Ks players with ideologically non-rigid Ss players does not prevent the rift from opening . Ss players , however , mend the rift after a Kb donor cooperates with an Ad recipient by assigning a favorable reputation to this donor . Liberal attitude towards moral consensus thus makes it impossible for social parasites to invade the inner circle ( Fig . 3a ) and can be seen as a layer of stability for the social structure . To confirm the robustness of the described mechanisms , we performed simulations ( results not shown ) with two additional sets of possible player types: and . It turned out that only the ideologically rigid combination of Ks and Kb players was vulnerable to an invasion by Ad , agreeing with the notion that liberal attitude towards moral consensus had a stabilizing effect on the society . In the context of the model robustness , we did not simulate unconditional cooperators nor the first-order scoring rule because it was shown that neither could maintain stable cooperation [8]; all else being equal , the former got eliminated in the presence of Sugden , Kandori , and Ad , whereas the latter , if not eliminated , became indistinguishable from Ad . The presence of unconditional cooperators , nonetheless , might have favored Kandori over Sugden to a certain extent because the Sugden rule would have encouraged more cooperativeness and hence higher costs in comparison with the Kandori rule [8] . In the modern world , two omnipresent processes affecting barriers to cooperation are the technological development ( lowering ) and globalization ( lowering ) . Because both of these processes make cooperation easier , our results imply ( to the extent game theoretical representations are valid in a complex reality ) that the modern world is conducive of ideologically non-rigid societies with presumably an increasing number of functioning democracies and more economic liberties [15] . Support can be found in indisputable growth of electoral democracy among the world's nations , especially over the past three decades , although the Democracy Index compiled by the Economist Intelligence Unit suggests that the overall quality of democracy is stagnating since the financial crisis of 2007–2008 . As for economic liberties , the average Economic Freedom of the World Index reported by the Fraser Institute indicates steady increase from 1980 until 2006 , but again a period of stagnation during the 2008–2012 global recession . Looking at our results from a different angle , a remarkable implication is that restrictive or protectionist policies aid the creation of ideologically rigid societies . Perhaps then it is not surprising that the above indices are stagnating in the midst of a five-years long recession . This is , after all , the second worst economic contraction since the Great Depression of the 1930s , which itself brought on a number of restrictive or protectionist policies , coinciding with the rise of multiple totalitarian regimes and ending only after the deadliest conflict in the human history . Analyzing the social role of ideological rigidity within the indirect reciprocity framework , we uncovered evolutionary outcomes that warn against restrictive or protectionist government policies . Yet to prevent from falling into the trap of naive rationalism or worse interventionism , a constant remainder is needed that these outcomes follow from a mathematically tractable representation of immensely complex human concepts such as value systems . We , therefore , perceive the present and similar studies as theoretical constructs that identify the potential , rather than the actual , drivers of social phenomena . Keeping this important distinction in mind does not diminish the multitude of opportunities for the field . Our understanding of the factors that promote ( e . g . punishment ) , disrupt ( e . g . corruption ) or shape the nature ( e . g . spontaneous in-group favoritism ) of cooperation is still quite limited .
The reputation dynamics control intra-generational partitioning of players according to their reputation . Intuitively , the outcomes of the reputation dynamics specify probabilities that the generation of players of a particular type will be assigned a particular reputation . More formally , we are concerned with a discrete probability measure defined on a sample set , , where the sample set is built from two basic constituents , the set of player types , , and the set of all possible reputations , . Because five distinct player types are considered , the set could generally be any combination of . It is beneficial , however , to display the results of extensive numerical simulations on ternary plots by referring only to the most illustrative outcomes . The main reason is that handling three player types at once permits us to effectively visualize and compare the results . The set is accordingly limited to 3-combinations with repetitions ( or 3-multisets ) of the set , where F and U denote a favorable and an unfavorable reputation , respectively . Hence , . With these basic constituents , the sample set is given by the Cartesian product of the form , so that implies . The discrete probability measure of concern , denoted by because it is closely related to the so-called honor score [16] , is fully defined by specifying how it operates on the elements of . To emphasize the dependence on the type and reputation of players , we introduce a short-hand notation . The letters and are used throughout the text to denote player types , primarily that of a donor and a recipient , respectively . The letters and are reserved for the type of observers and outsiders , respectively . It is also useful to reserve the letters and for the reputation of donors and recipients , respectively . Note that and , therefore , stand for three different reputations . When interested only in the reputation from the viewpoint of type observers , we can write and . In addition , to keep formulas for the probabilities more tractable , it is helpful to introduce two auxiliary functions as shown henceforth . The first of the two auxiliary functions , denoted , is called the action rule . Because each interaction in the game involves two players , a donor and a recipient , the action rule specifies the probability of an action being undertaken by the type donor towards the recipient with the reputation from the donor's viewpoint . Only two actions are possible , so that , where C and D stand for cooperation and defection , respectively . It follows , . For the action rule isand . Ad players never cooperate , thus . The second auxiliary function , denoted , is called the assessment rule . Because observers assign new reputations to donors after every interaction , the assessment rule represents the probability that the donor will be assigned the reputation by the type observer if the recipient's reputation from the observer's viewpoint is and an action is taken . Consequently , . The case is presented in Table 1 , where the distinction between and formally defines Sugden and Kandori observers . Similarly with the action rule , . Ad players treat every donor unfavorably , thus . By combining action and assessment rules , we can formally express how action-assessment strategies determine the probabilities that the donor will be assigned the reputation conditional on all relevant circumstances . Such conditional probabilities are crucial for calculating the probabilities . If the recipient is an insider , the relevant circumstances are specified by the type of observer ( ) and donor ( ) , as well as the recipient's reputation from observer's ( ) and donor's ( ) viewpoints . Accordingly , we introduce ( 1 ) If the recipient is an outsider , both action and assessment differ between shonin and bushi players , thus providing a way to formalize the distinction between the two . The former players make an effort to learn the outsider's reputation , whereas the latter simply dismiss the outsider as a player with an unfavorable reputation . Such a situation implies that outside recipients , who are by assumption of type , are perceived favorably by the observer and the donor , i . e . , if and only if . Consequently , we can introduce the probability as an analogue to conditional only on the type of observer ( ) , donor ( ) , and outsider ( ) by ( 2 ) We mentioned that the reputation dynamics controlled partitioning of players according to their reputation within a generation . Therefore , each generation plays many rounds of the game , whereby every player serves both as a donor and as a recipient once per round . When serving as a donor , the player encounters either an insider recipient with probability or an outsider recipient with probability . If the recipient is an insider , the probability of it being a player of type is and the probability of its reputation being is , where denotes the current round of the game . Using Eqs . ( 1 ) and ( 2 ) , the probability that the type donor is assigned the reputation for the next round , , becomes ( 3 ) After ( infinitely ) many rounds of the game , the probabilities converge to the equilibrium values defined by . These equilibrium values are then used to simulate the evolutionary dynamics of value systems inside the inner circle . The evolutionary dynamics of value systems inside the inner circle is modeled using the replicator equations . If we denote the fraction of the type players at the generational time with , so that , then the same fraction in the next generation , , is given by ( 4 ) where is the fitness of the type players and is the average fitness ( both at time ) . Fitness is a function of the equilibrium probabilities ( and the parameters , , , and ) because players of a given type receive the payoff to the extent they are perceived favorably by their respective donors and incur the cost to the extent they perceive their recipients favorably . Assuming that outsiders are of type , i . e . , the above considerations can be written in general mathematical terms aswhere is an arbitrary basic level of fitness and is the usual Kronecker delta symbol . Besides , two additional summands appear in the last equation . The first of the two summands represents the difference between the benefits received and the costs incurred from within-group encounters . Note that the benefits from type players are received only if the reputation is favorable , i . e . . Similarly , when encountering type players the costs are incurred only if the reputation is favorable , i . e . . The second of the two summands also represents the difference between benefits and costs , but now as a result of cross-border encounters . Here , insiders who are of the same type as outsiders , i . e . when , receive the benefits if their reputation is favorable ( ) and incur the cost with certainty because the outside world is assumed to be in a stable equilibrium populated by the type players . Social parasites behave opportunistically in the sense that they receive the benefits whenever their reputation is favorable from the viewpoint of the type observer . We performed numerical simulations based on the described methodology to ( i ) visualize convergence of the model over the generational time scale , ( ii ) delineate the domains of attraction , and ( iii ) estimate their sizes . We achieved these goals in several steps . First , we defined a grid with 5050 points distributed uniformly over the ternary domain , , and . Coordinates of each grid point served as the initial conditions for one model run . Every run consisted of many generational time steps , where in a single step multiple rounds of the game played out according to Eq . ( 3 ) . Instead of presetting the number of rounds , we waited until the difference reached the desired accuracy . The resulting approximation of the equilibrium probabilities allowed us to calculate the fitness of all player types and advance their respective fractions , , into the next generation , , using Eq . ( 4 ) . Eventually , the model converged to one of the attractors , forming a link between the starting point and the attractor . In the second step , we chose a uniform subset ( 134 points ) of the initial grid for which sample paths over the generational time were stored and subsequently visualized in the ternary plots ( Figs . 1 and 3 ) . For the visualization , we used curved arrows to characterize the direction and the rate of convergence along sample paths; the longer an arrow , the faster the convergence rate along that particular path . The third step began after completing all 5050 runs for a fixed parameter set . Because every grid point had been linked with an attractor , we could isolate the neighboring points that belonged to two different domains of attraction . Segments connecting such neighboring points were further subdivided with five equidistant points to provide the initial conditions for extra runs in which an even closer pair belonging to two different domains of attraction could be determined . The process continued until the distance between the neighboring points reached the desired accuracy and thus revealed the location of the border between the adjacent domains of attraction . In the final step , we calculated the fraction of grid points linked to each attractor as an estimate of the size of the corresponding domain of attraction . | Attitudes , beliefs , and resulting value systems may represent important motivational and decision-making factors that have strong impact on cooperation in a society . Accordingly , understanding the social function of value systems is a topic of great interest in evolutionary biology , but one where progress is made difficult by the sheer complexity of values-inspired behaviors . Here , we argue that considerable theoretical progress can be made within the indirect reciprocity framework . We show in the context of indirect reciprocity how to construct stylized value systems from a mathematically formalized notion of ideological rigidity . Our simulations indicate that politically imposed restrictions and protectionism favor the evolution of ideologically rigid value systems . The complete collapse of cooperation also arises as a possible evolutionary outcome . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Methods"
] | [
"social",
"systems",
"social",
"sciences",
"evolutionary",
"modeling",
"biology",
"and",
"life",
"sciences",
"sociology",
"computational",
"sociology",
"evolutionary",
"theory",
"evolutionary",
"biology",
"social",
"research",
"computational",
"biology"
] | 2014 | Barriers to Cooperation Aid Ideological Rigidity and Threaten Societal Collapse |
The current evidence-base for recommendations on the treatment of cutaneous leishmaniasis ( CL ) is generally weak . Systematic reviews have pointed to a general lack of standardization of methods for the conduct and analysis of clinical trials of CL , compounded with poor overall quality of several trials . For CL , there is a specific need for methodologies which can be applied generally , while allowing the flexibility needed to cover the diverse forms of the disease . This paper intends to provide clinical investigators with guidance for the design , conduct , analysis and report of clinical trials of treatments for CL , including the definition of measurable , reproducible and clinically-meaningful outcomes . Having unified criteria will help strengthen evidence , optimize investments , and enhance the capacity for high-quality trials . The limited resources available for CL have to be concentrated in clinical studies of excellence that meet international quality standards .
It is important to harmonize and improve clinical trial methodology for cutaneous leishmaniasis ( CL ) ; currently , treatment options are few and the quality of the supporting evidence is generally inadequate , making the strength of recommendations for the treatment of this disease inadequate . To improve on the case management and control of CL , better treatment modalities with reliable evidence of the efficacy , safety , tolerability and effectiveness is required . High-quality clinical trials are essential to determine which therapeutic interventions can confidently be recommended for treating which form of CL . Today , this is unfortunately not the case in numerous instances . The inadequacies of trials of different treatments of CL has been documented by two WHO-supported Cochrane systematic reviews [1] , [2] which included 97 randomized controlled trials on treatments for Old World and American CL . They revealed critical issues related to the methodological quality of the design and reporting of these clinical trials , which make it difficult to compare results , meta-analyse the studies , and draw generalizable conclusions . Weaknesses ranged from the inadequacy of study design ( including appropriate controls , endpoints , outcome measures , follow-up times ) , execution ( randomization , allocation concealment , blinding ) , analyses and reporting ( e . g . use of disparate endpoints ) [3] . They also found a large number of trials that did not meet basic criteria , and could not be included in the analyses . This makes a highly compelling and cogent case for defining and harmonizing elements related to the design , conduct , analysis , clinical relevance , and reporting of trials , and ultimately study acquiescence by regulatory agencies . Improving the quality of studies and harmonizing protocols will make meta-analysis more informative and thus strengthen evidence for recommendations on treatment and case management . Furthermore , conducting inadequate trials may lead to inappropriate conclusion , is both unethical and an inefficient use of the limited resources available for research into this neglected disease . As heterogeneity is an inherent feature of CL ( reflecting the variety of species and manifestations ) , there are obvious challenges in designing and interpreting trials to assess interventions for CL which will allow deriving generalizable results and recommendations . The objective of this paper is twofold: This paper focusses on CL trial-specific issues; it only touches upon more general aspects of clinical trial conduct , which are extensively addressed in a number of relevant papers and documents . For instance the Global Health Trials website [4] offers several resources including a trial protocol tool [5] . The collective name of CL comprises several manifestations caused by different Leishmania species in the Old and the New World ( OWCL and NWCL ) and clinical trial methodology should be adapted to this spectrum of conditions . CL is caused by organisms of the L . mexicana complex and Viannia sub-genus ( L . braziliensis and L . guyanensis complex ) in the New World and L . major , L . tropica and L . aethiopica in the Old World . L . infantum in both Worlds and L . donovani in the Old World can also cause CL . The wide spectrum of clinical manifestations , natural histories and responses to treatment observed in CL patients is accounted for by the combination of parasite's intrinsic differences and patient's genetic diversity . The time required for natural cure ( “self-healing” ) is poorly defined and varies widely; it is generally accepted that lesions caused by L . mexicana in the New World and L . major in the Old World heal spontaneously in a time varying from a few weeks to several months in the majority of patients – except new foci ( where the disease tends to be aggressive and self-healing is uncommon ) , and as opposed to other species ( where spontaneous healing barely occurs or requires years ) . Bacterial super-infections are also frequent and can interfere with healing . The natural history of the disease must be accounted for when designing a clinical trial . Good knowledge of the disease characteristics at the trial site is essential; it is not possible to extract generalizable data from the published literature . For instance , when considering the placebo arms of randomised controlled trials ( RCTs ) from the Cochrane systematic review of OWCL1 , 3-month cure rates for L . major were 21% in Saudi Arabia and 53% in Iran with oral placebo . With a topical placebo , they varied from 13% to 63% at 2 months in Iran and were 61% in Tunisia at 2 . 5 months . For L . tropica , cure rates were 0%–10% with oral placebo . In the New World , the information is scarce and more variable , ranging from 0% cure rate at one month in Panama [6] to 37% at 12 months in Colombia [7] for lesions most probably caused by L . panamensis . In Guatemala , using topical or oral placebos a 68% cure rate was reported at 3 months for lesions due to L . mexicana and only 2% for lesions due to L . braziliensis [8] , while other studies have reported cure rates of 27% and 39% in the general population at 3 and 12 months respectively [9] , [10] . In Ecuador , in a small group of 15 patients , a cure rate of 75% at 1 . 5 months ( no speciation but likely L . panamensis ) was reported without any treatment [11] . The examples above illustrate the need to acquire and factor in local data on the natural history of disease in order to assess more accurately treatment performance . A wide variety of treatment modalities has been reported for CL , but none has been shown to be universally effective . Treatment response varies according to a range of factors , including the Leishmania species , the patient immune status and age , the number and localization of the lesions , the severity of the disease , the treatment given and the route of administration , etc . Treatment would benefit both the individual patient but also reduce the burden of human reservoirs in the case of anthroponotic CL , and prevent super-infection and the resulting complications . The choice of treatment , either local or systemic , is usually based on the size , number and localization of lesions , lymphatic spread or dissemination , patient's immune status , cost , risk-benefit and the availability of the treatment itself in the country . Currently available treatment options ( systemic and topical ) can be found in the WHO 2010 technical report [12] .
One of more of the following criteria may apply . Table 1 illustrates how to apply the different entry criteria based on the type of study ( Phase 2–4 ) and treatment being tested ( systemic or topical ) . The final decision , however , must be taken based on prior knowledge accrued during the pre-clinical and phase I studies .
The assessment and reporting of the safety , toxicity and tolerability of treatments , while an essential component of the evaluation , is often overlooked in CL clinical trials . Topical treatments may produce local events at the site of the lesion ( like irritation ) ; systemic treatments may cause generalised signs or symptoms , including changes in laboratory values . Events should be reported and graded using standard nomenclature and criteria of severity . Whenever possible , events must be combined under a syndrome or diagnosis . It is important to comply with regulations for filing serious events; specific requirements exist for timely reporting accoriding to national regulations ( health authorities , regulatory authorities , ethics committees ) . However , investigators must be alerted to the fact that definitions and rules for reporting may evolve with time and are not fully harmonised between countries .
All trials should be registered ( see: the WHO International Clinical Trials Registration Platform ( WHO-ICTRP ) and reported , whether the results are favourable , unfavourable or inconclusive – both for ethical and scientific reasons . Traditionally , the importance of negative results has been underestimated both by researchers and publishers; publishing only positive results will bias knowledge . The CONSORT checklist ( study design , analysis and interpretation ) and flow diagram ( patient attrition throughout the study ) should be followed [63] . All major journals today do not publish papers on trials that have not been registered and do not follow the CONSORT guidelines ( see example in Figure 9 ) . The protocol must be clear as to the population for analysis – typically: intent-to-treat ( ITT ) , modified ITT ( mITT ) and per-protocol ( PP ) . The basis for exclusion of patients from the analysis must be provided . Patients withdrawn because they could not tolerate treatment or because they required rescue treatment must be accounted for . The analytical plan should be finalised before freezing the data for analysis . Like any other trial , an appropriate data management process is critical in order to have high-quality data , statistical analyses and results . For this purpose , the data management software adopted must provide a secure location for the clinical data , user rights and profiles along with password protection , as well as an audit trail . Capacity for data management is often scarce in CL-endemic countries , including both the availability of appropriate software with auditable track , and trained data managers . In these countries there is also a general shortage of statisticians to help design and to analyse and report on trials . Capacity building efforts should be organized to increase competences of research teams in this important area .
Clinical trials must be conducted in accordance with current international standards of Good Clinical Practice ( GCP ) , an international ethical and scientific quality standard for designing , conducting , recording and reporting trials that involve the participation of human subjects . Compliance with this standard provides public assurance that the rights , safety and well-being of trial subjects are protected , consistent with the principles that have their origin in the Declaration of Helsinki , and that the clinical trial data are credible . When GCP standards are followed , the quality of data from clinical trials is adequate to make informed clinical and policy decisions . There is a belief among some that GCP guidelines are only for “registration” studies and not for all clinical trials . However , the principles of GCP should be applied to all clinical studies with any intervention conducted at any stage of development that may have an impact on the safety and well-being of human subjects . Implementation of GCP procedures requires initial training and practice and is best served when trial personnel at a site accept and understand a culture of GCP . Maintaining a GCP environment requires constant training and reinforcement and is a process that requires continuous growth in a site and personnel . Accepted GCP standards include those published by the International Conference on Harmonization ( ICH ) and the World Health Organization ( WHO ) . The ICH GCP guideline is published under Efficacy ( E6 ) and is often referred to as ICH E6 GCP guideline [64] . A summary review of the principles of GCP are found in the WHO handbook [65] . At the same time , it should be clear that GCP is not about dogma , but rather patient's care and reliability of data , and that the context within which trials occur should be accounted for . A proper balance between the goals of the clinical study and the documentation required has been proposed [66] . The amount of written documentation and the degree of detail required by GCP procedures can be a shock to investigators not used to working in this environment . Although the conduct of clinical trials under GCP with external monitors and proper data management will inevitably increase the cost of studies , it is imperative that higher quality studies in CL be conducted . For all trials involving human subjects , ethics review and approval must be sought from appropriate boards/committees at the institution ( local and/or international ) and/or country level as required . It is imperative that all clinical studies are conducted in accordance to the international and country regulations and laws .
The opinions expressed in this paper are those of the authors; the authors alone are responsible for the views expressed in this publication and they do not necessarily represent the decisions , policy or views of the WHO . Material has been reviewed by the Walter Reed Army Institute of Research . There is no objection to its presentation and/or publication . The opinions or assertions contained herein are the private views of the author , and are not to be construed as official , or as reflecting true views of the Department of the Army or the Department of Defense . | Solid evidence is needed to decide how to treat conditions . In the case of cutaneous leishmaniasis , the diversity of clinical conditions , combined with the heterogeneity and weaknesses of the methodologies used in clinical trials , make it difficult to derive robust conclusions as to which treatments should be used . There also other imperatives - ethical ( not exposing patients to treatments that cannot be assessed adequately ) and financial ( optimize use of limited resources for a neglected condition ) . This paper is meant to provide clinical investigators with guidance for the design , conduct , analysis and report of clinical trials to assess the efficacy and safety of treatments of this condition . | [
"Abstract",
"Introduction",
"Defining",
"trial",
"participants",
"Endpoint",
"-",
"outcome",
"measures",
"and",
"therapeutic",
"assessment",
"Study",
"registration",
"and",
"reporting",
"Complying",
"with",
"regulations",
"Disclaimer"
] | [
"medicine",
"infectious",
"diseases",
"leishmaniasis",
"neglected",
"tropical",
"diseases"
] | 2013 | Methodology of Clinical Trials Aimed at Assessing Interventions for Cutaneous Leishmaniasis |
One of the hallmarks of viral immune evasion is the capacity to disrupt major histocompatibility complex class I ( MHCI ) antigen presentation to evade T-cell detection . Cowpox virus encoded protein CPXV203 blocks MHCI surface expression by exploiting the KDEL-receptor recycling pathway , and here we show that CPXV203 directly binds a wide array of fully assembled MHCI proteins , both classical and non-classical . Further , the stability of CPXV203/MHCI complexes is highly pH dependent , with dramatically increased affinities at the lower pH of the Golgi relative to the endoplasmic reticulum ( ER ) . Crystallographic studies reveal that CPXV203 adopts a beta-sandwich fold similar to poxvirus chemokine binding proteins , and binds the same highly conserved MHCI determinants located under the peptide-binding platform that tapasin , CD8 , and natural killer ( NK ) -receptors engage . Mutagenesis of the CPXV203/MHCI interface identified the importance of two CPXV203 His residues that confer low pH stabilization of the complex and are critical to ER retrieval of MHCI . These studies clarify mechanistically how CPXV203 coordinates with other cowpox proteins to thwart antigen presentation .
Detection of viral infection by CD8 T cells relies on major histocompatibility complex class I ( MHCI ) presentation of virally derived peptides at the cell surface . Not surprisingly , a wide variety of viruses have evolved mechanisms to disrupt antigen presentation by targeting the assembly and trafficking pathways used by MHCI proteins [1] , [2] . The most common immune evasion mechanism appears to be blockade of cytosol-to-endoplasmic reticulum ( ER ) peptide transport by the transporter associated with antigen processing ( TAP ) [3]–[10] . However , other viruses target molecular chaperones to impair the quality of peptide loading without curtailing peptide supply [11] , [12] . The quality of peptide loading by MHCI is initially controlled by the peptide loading complex ( PLC ) made up of TAP , tapasin ( Tpn ) , ERp57 , and calreticulin ( CRT ) [13] . Prior to binding an optimal peptide , the PLC retains in the ER nascent MHCI heavy chains ( HCs ) assembled with beta-2 microglobulin ( β2m ) . Within the PLC , the MHCI-dedicated chaperone Tpn bridges the HC/β2m complex with TAP . Once a peptide of suitable affinity binds to the HC/β2m complex , the fully assembled MHCI is released from the PLC to transit to the cell surface; and perhaps not surprisingly , there are examples of viral immune evasion proteins that impair peptide loading by blocking PLC assembly [11] , [12] . In addition to PLC-imposed quality control , non-PLC-associated CRT uses a KDEL-dependent mechanism to retrieve suboptimally loaded MHCI from the early Golgi to the ER to improve peptide binding [14] . This ER retrieval is dependent upon the C-terminal KDEL sequence of CRT that confers binding to the KDEL receptor ( KDELR ) in the early Golgi in a pH-dependent manner [15] . Several viral immune evasion proteins appear to directly target MHCI proteins , but only adenovirus ( AdV ) E3-19K and human cytomegalovirus ( HCMV ) US2 have been shown to directly bind MHCI luminal domains [16] , [17] . E3-19K impairs MHCI egress from the ER by either an ER-retention mechanism dependent on its cytoplasmic tail [18] or its ability to prevent Tpn bridging MHCI to TAP [11] , while US2 targets MHCI for ER-associated degradation ( ERAD ) [19] . E3-19K and US2 both exhibit distinct class Ia allele preferences [20]–[22] that may help these viruses evade natural killer ( NK ) cell cytotoxicity on the basis of missing self [23] . Alternatively , viruses may encode separate proteins to undermine NK cell surveillance [24] . Interestingly , E3-19K has also been reported to target various MHCI assembly intermediates , and mutagenesis analyses suggest that E3-19K may interact with an MHCI surface similar to that bound by US2 [20] , [22] . The only structural study of direct MHCI sabotage revealed that US2 uses an Ig-like fold to bind under the MHCI-binding platform near where the N-terminus of the peptide is anchored [25] . Although US2 was crystallized bound to fully assembled MHCI , cellular studies suggest US2 also targets HC before full assembly with peptide and/or β2m [26] . In any case , the structural basis for how US2 , E3-19K , or any other viral immune evasion protein discriminates MHCI alleles and/or assembly intermediates has not been previously reported . While many viruses exhibit strict host specificity , some orthopoxviruses are able to productively infect a wide variety of mammalian species and encode an array of immunomodulatory genes that target both cell intrinsic and extrinsic antiviral responses [27] . Yet until recently , orthopoxviruses were not known to target antigen presentation . The orthopoxvirus cowpox ( CPXV ) expresses two unrelated immune evasion proteins , CPXV012 and CPXV203 ( UniProt [UNP]: Q8QMP2 ) , which use different mechanisms to block antigen presentation in both human and murine cells [28]–[30] . CPXV012 is a small type II transmembrane protein that blocks peptide transport by TAP [29] , [30] . By contrast , CPXV203 is a soluble protein that prevents MHCI proteins from trafficking to the plasma membrane by a mechanism dependent upon its C-terminal KTEL sequence , a motif recognized by the KDELR [28] . To initially probe binding partners , Byun et al . ( 2007 ) showed that CPXV203 co-precipitated with MHCI and not TAP . These findings implied that CPXV203 binds MHCI lumenal domains or an associated molecule before and/or after peptide assembly [28] . Furthermore , CPXV203 was found to down regulate MHCI proteins in both murine and human cell lines during normal poxvirus infection [29] , [30] . This ability to broadly inhibit MHCI by CPXV203 may help explain productive CPXV zoonotic infection of various mammalian species other than small rodents , the apparent CPXV host reservoir [27] . Indeed , mutant cowpox viruses lacking both CPXV012 and CPXV203 demonstrate attenuated virulence in a cytotoxic T lymphocyte ( CTL ) -dependent manner [29] , in contrast to other viral proteins that target MHCI that do not appear to significantly modulate primary infection in vivo [31] , [32] . Here we provide a precise understanding of how CPXV203 binds to a broad array of MHCI complexes that includes both classical and non-classical molecules . Biosensor studies indicate that CPXV203 binds MHCI weakly at the pH found in the ER , but the affinity and half-life are significantly augmented at the more acidic conditions found in the Golgi . Crystallographic analysis reveals that CPXV203 adopts a β-sandwich topology reminiscent of poxvirus chemokine-binding proteins , and this domain engages evolutionarily conserved MHCI determinants available only on fully assembled MHCI . We also undertook mutagenesis analysis that supports the structural model and uncovered the critical functional role played by two CPXV203 His residues in the pH regulation of complex stability . Together these data suggest that CPXV203 works cooperatively with the endogenous KDEL-mediated Golgi retrieval process to promiscuously target fully assembled MHCI , thereby preventing T-cell killing of cowpox infected cells .
To ascertain which MHCI assembly state ( s ) is targeted by CPXV203 , association with the HC of murine H-2Kb ( UNP: P01901 ) was monitored by co-precipitation in wild-type and β2m-deficient cells . CPXV203 only co-precipitated with Kb HC in cells expressing β2m ( UNP: Q91XJ8 ) ( Figure 1A ) , suggesting that heterodimer assembly is required for CPXV203/MHCI association . To further assess whether this association was dependent upon the PLC , CPXV203 was expressed by transduction in cells lacking either TAP or Tpn , which present low levels of fully assembled MHCI . As shown in Figure 1B , CPXV203 dramatically reduced MHCI surface expression in cells lacking TAP or Tpn , whereas the TAP inhibitor CPXV012 did not affect surface expression in these PLC-component deficient cells . We also found that CPXV203 comparably downregulates MHCI expression in cells with and without CRT ( Figure 1C ) , suggesting that CPXV203 expression does not grossly disrupt CRT-associated ER quality control as could potentially occur through KDELR competition . In further support of this conclusion , CPXV203 does not interfere with PLC assembly , as shown by normal TAP/Tpn association and normal steady-state levels of CRT ( Figure 1D ) . Previous studies found comparable peptide loading in cells with and without CPXV203 , and failed to identify association of CPXV203 with the PLC [28] . Taken together , these findings provide compelling evidence that CPXV203 regulates the surface expression of fully assembled MHCI after dissociation from the PLC without impairing PLC function . We next sought to examine whether CPXV203 directly binds to MHCI using soluble recombinant proteins in biophysical assays . We observed that CPXV203 binds Kb with an affinity of KD , Kin = 480 nM at pHER 7 . 4 using surface-plasmon resonance ( SPR ) ( Figure 2A ) . The expansion of these studies to additional MHCI molecules revealed that CPXV203 exhibits low affinity interactions ( KD , Kin = 82–10 , 500 nM ) with five different murine Ia alleles ( Db , Dq , Kb , Kd , Ld ) and a primate allele ( Ceat-B*12 ) ( Table S1 ) . We also examined a non-classical MHC Ib protein , murine thymic leukemia tumor antigen or TL ( T3b ) , which was engaged by CPXV203 with similar affinity and kinetics at pHER 7 . 4 as observed for Kb . Unlike classical MHCI proteins that require peptide loading to assemble , TL pairs with β2m and is stable in the absence of ligand binding . Thus , it appears that the requirement for peptide binding to classical MHC Ia proteins for CPXV203 engagement is based on the role peptide loading plays in assembly and stability rather than direct recognition . Promiscuous CPXV203/MHCI association fits well with the previously published data that CPXV203 downregulates murine H-2D and -K alleles , though the affinities were weaker than those previously reported for the viral ER retention protein E3-19K ( 11–18 nM ) [22] . The weaker than expected affinity of CPXV203 for MHCI led us to evaluate a variety of buffer conditions that might more closely reproduce ER/Golgi conditions ( divalent cations: Ca2+ , Mg2+ , Zn2+; ATP; pH 6–8 ) . Of these changes , only low pH augmented CPXV203/MHCI affinity with a decrease to pHGolgi 6 . 0 increasing the affinity ∼50-fold ( KD , Kin = 10 nM , Figure 2B ) . This striking enhancement occurs through both an increased on-rate ( ka ) and a decreased off-rate ( kd ) for all tested murine and primate alleles ( Table S1 ) . We confirmed these results using a separate biophysical technique , biolayer interferometry ( BLI ) , where the equilibrium response of CPXV203 binding murine ( H-2Dk , -Kb , k , -Ld; TL ) and primate MHCI ( Mamu-A*01 , Patr-B*0802 ) was monitored as a function of pH ( 7 . 6–6 . 0 ) ( Figure 2C ) . These binding studies demonstrate that the stability of CPXV203/MHCI complexes is pH regulated to favor association in the Golgi rather than the ER . Importantly , similar observations have been made for the binding of KDEL bearing ligands to the KDELR [15] . To address whether enhanced binding of CPXV203 to MHCI at low pH results from changes in the stoichiometry of the complex , multi-angle light scattering ( MALS ) experiments were undertaken that demonstrated a 1∶1 stoichiometry that was insensitive to pH manipulation from pH 6 . 5–8 . 5 ( Figure S1 ) . We also undertook circular dichroism spectra analysis that indicated that the conformation of these proteins ( alone or in complex ) does not change significantly as a function of pH ( unpublished data ) . These results support the 1∶1 binding model used in our biosensor analysis and suggest that CPXV203/MHCI pH regulation likely involves only small local effects . We next pursued crystallographic studies of CPXV203 in complex with MHCI to better understand the nature of the interaction . Utilizing the observation that CPXV203 binding affinity increases with decreasing pH , we crystallized SeMet-labeled CPXV203 in complex with Kb loaded with SIINFEKL ( OVA257–264 ) at pH 5 . 55 and determined the structure at 3 . 0 Å resolution ( Figure 3A; Table S2 ) . Initial molecular replacement phases using MHCI alone were greatly improved through cross-crystal averaging ( Figure S2A–S2D ) [33] , which allowed a preliminary backbone trace of CPXV203 to be built . Subsequently , molecular replacement-single-wavelength anomalous dispersion ( MR-SAD ) was used to identify eight SeMet sites and introduce anomalous phase information ( figure of merit [FOM] 0 . 604 ) that improved map quality to the point where the complete CPXV203/MHCI complex could be built and refined . The structure reveals that CPXV203 binds below the MHCI peptide-binding platform , contacting both the HC ( α2- and α3-domains ) and β2m . Comparison of Kb free and bound by CPXV203 did not indicate any significant changes associated with viral protein engagement . The general footprint located below the α2-1 helix of Kb is supported by serological experiments whereby we determined whether CPXV203 competed with monoclonal antibodies specific for well-characterized epitopes . Direct binding competition was observed for two monoclonal antibodies ( MAbs ) ( AF6-88 . 5 . 3 and Y-3 ) that have been mapped precisely to this region , and no competition was observed for three MAbs mapped outside of the CPXV203 footprint ( Figure S2F–S2H ) . We note that while HCMV US2 also binds MHCI below the peptide-binding platform , the CPXV203 footprint is completely distinct and , strikingly , overlaps with that of Tpn , CD8 , and NK cell receptors ( NKRs ) ( Figure 3B ) . The structure of CPXV203 does not resemble any structurally characterized viral or host protein known to interact with MHCI . The single domain of CPXV203 ( 209 aa ) is a globular β-sandwich that is stabilized by five disulfide bonds conserved in all T4 poxvirus protein family members ( Figures 4A , 4B , S3A ) . The core β-sandwich consists of two parallel β-sheets ( β-sheet I: β1 , β5 , β6 , β10; β-sheet II: β2 , β3 , β4 , β7 , β8 , β9 ) made up of anti-parallel strands with one parallel strand interaction ( β7/β9 ) bridging the two segments of β-sheet II ( Figure 4B ) . Three of these disulfide bonds appear to stabilize the h4-loop-h5 arrangement used to engage the MHCI α2-domain . A search for structurally similar proteins indicates that the CPXV203 β-sandwich core resembles the structurally characterized poxvirus chemokine binding proteins ( CKBPs ) , such as the vCCI-like protein encoded by ectromelia virus , EVM001 [34] , which exhibits an RMSD of 3 . 0 Å for 143 aligned residues ( Figure S3B; Table S3 ) . CPXV203 and the poxvirus CKBPs engage their ligands using completely distinct binding surfaces located on opposite faces of the β-sandwich core ( Figure 4C ) . While vCCI-like proteins use β-sheet II to bind chemokines , CPXV203 primarily uses β-sheet I elements . Interestingly , the Ectromelia virus CrmD-CTD ( SECRET domain ) also appears to use β-sheet I elements to bind chemokines , and it shares with CPXV203 a distinct β7–β9 junction relative to vCCI-like proteins that increases β-sheet I accessibility through the conversion of a flexible loop into a β-sheet II strand ( Figure 4D ) . CPXV203 has further differences with the vCCI core with the replacement of vCCI β13-β14 with two α-helices ( h4 and h5 ) , a modification that also exposes CPXV203 β10 to interact with β2m . For CPXV203 , these topological changes remove potential steric clashes ( Figure S3C ) , increase solvent-accessibility of the conserved β5–β6 loop ( source of nearly all α3 contacts ) , and create the primary sources for both α2 ( h4–h5 ) and β2m ( β8 and β10 ) contacts ( Figure 4B; Table S4 ) . Thus , while CPXV203 is clearly structurally related to poxvirus CKBPs , significant modifications are clearly evident that uniquely allow it to recognize MHCI . To understand the structural basis of how CPXV203 interacts with such diverse MHCI-family proteins , we analyzed the conservation of MHCI contacts and the similarity of these contacts to those used by other MHCI-binding proteins . CPXV203 promiscuously binds MHCI through a large , somewhat nonpolar interface divided into three distinct contact regions ( α2 , α3 , and β2m domains ) ( Figures 5A , 5B , S4A , S4B ) . The arrangement of these contact regions is only available in fully assembled MHCI , and as such CPXV203 binds MHCI in an assembly-dependent manner that is peptide-independent as long as MHCI assembly is also peptide-independent , as is the case for TL . Comparison of the CPXV203/MHCI interface to similar interfaces reveals that the total buried surface area ( BSA ) is significantly larger than most other complexes , CPXV203 buries >200 Å2 more main-chain ( MC ) than any similar MHCI-binder , and only CPXV203 divides its interface nearly equally among the platform ( α1/α2 ) , β2m , and α3 ( Figure 5A; Table S5 ) . CPXV203 recognizes MHCI elements that are extremely well conserved in all murine pMHCI: overall , 86%; CPXV203 contacts , 91%; CPXV203 side chain ( SC ) contacts , 95%; five invariant SC contacts . Further , CPXV203 recognizes core structural features of the MHCI fold by anchoring each of the three domain interfaces through a buried MC-MC hydrogen bond and two to three MC-SC hydrogen bonds ( Figure S4C–S4E; Table S4 ) . Through these contacts , CPXV203 recognizes seven backbone positions conserved by the MHCI fold and coordinates conserved MHCI side chains within the α3 interface ( Q226 , D227 , E229 ) also required for Tpn and CD8 association [35]–[37] . Further , the presence of CPXV203 His residues opposite negatively charged α3 domain residues ( H75-D227 , H80-E229 ) suggests these may be pH-regulated interactions , though only H80-E229 is close enough to form a direct contact ( 3 . 5 Å versus 7 . 3 Å ) . Finally , we have identified that CPXV203 downregulates the non-classical MHCI molecule H2-M3 , while mCD1d escapes CPXV203 retrieval ( Figure S4F , S4G ) . Our structural results support the idea that escape by mCD1d is facilitated in part by a charge reversal at position 229 ( mCD1d H233 – CPXV203 H80 ) and the orientation of mCD1d Q230 away from the interface due to an altered CD-loop conformation ( Figure S4H ) . This structural investigation explains promiscuous MHCI retrieval by CPXV203 , as it specifically targets a tri-domain interface of evolutionarily conserved contacts that would only be presented by fully assembled MHCI . We assessed the functional relevance of specific determinants within the CPXV203/MHCI binding interface by extensive mutagenesis of both CPXV203 and Kb . Mutants were assayed for loss of function by rescue of surface Kb expression or lack of physical association by co-immunoprecipitation ( co-IP ) . Single mutations in either Kb or CPXV203 from all three interaction sites ( Figure 5 ) were tested , but only α3 interface mutations Kb E229Y and CPXV203 F76A significantly rescued Kb surface expression ( Figure 6A , 6B ) . Furthermore , double mutations within the α3 interface ( Kb D227K/E229Y , CPXV203 H75A/H80A , Kb Q226A/CPXV203 F76A , Kb E229Y/CPXV203 H75A , Kb E229Y/CPXV203 F76A ) or the simultaneous mutation of interfaces α2 and α3 ( CPXV203 Y161A , F76A ) significantly enhanced Kb rescue , with some mutants displaying complete ablation of CPXV203 function ( Figure 6A–6C ) . Physical association ( CPXV203-HA/Kb ) was impaired more dramatically than Kb rescue by single α3 interface mutations ( Figure 6D ) , though it should be noted that the HA-tag might impair association . In any case , these experiments clearly demonstrate the functional importance of our structurally defined interface in CPXV203-mediated MHCI association and retrieval . To extend these findings , biosensor studies were undertaken to probe the contribution of individual interface residues in binding and pH regulation . Equilibrium analysis ( BLI , pHER 7 . 4 ) of CPXV203 and MHCI mutants further confirmed the three-site binding footprint ( Figure 5 ) and clearly distinguished the CPXV203 binding site from those of E3-19K , US2 , and MAb Y-3 ( Table S6 ) . Alanine mutation of several CPXV203 residues within the α3-domain interface ( including His residues 75 and 80 ) had a pronounced deficit in binding ( similar to co-IP ) . CPXV203 H75 and H80 were selected for kinetic analysis ( SPR ) based on their chemical environment ( Figure S4E ) and previously described sensitivity to alanine mutation ( functional and association ) . Alanine mutation of either His ablated the off-rate ( kd ) enhancement at low pH , while maintaining a similar on-rate ( ka ) enhancement ( Figure 6E; Table S7 ) . Thus at low pH , α3 interface His residues act to extend the CPXV203/MHCI half-life ( 6 s–73 s ) , while a separate interaction appears to regulate the faster association ( ka ) observed at low pH . Consistent with their importance , the double mutant ( H75A , H80A ) displayed extremely weak affinity that prohibited accurate kinetic analysis , though equilibrium BLI assays clearly support the greater functional deficit observed for this mutant ( Table S6 ) . These investigations indicate that CPXV203 engages MHCI through critical pH-regulated interactions with conserved MHCI α3-domain determinants , while the α2 and β2m domain interfaces may enable CPXV203 to bind fully assembled MHCI with broad specificity .
Viral infection of mammalian hosts can be greatly facilitated by viral proteins that confer the ability to evade CTL detection and clearance . Not surprisingly , viruses have evolved a wide variety of strategies to reduce cell surface presentation of viral peptides on MHCI [1] , [2] . The cellular , structural , and biophysical results reported here provide a complete picture of one such strategy , as CPXV203 was shown to directly bind fully assembled MHCI in a manner that is regulated via the normal pH gradient that exists between the ER and Golgi compartments ( Figure 7 ) . Though CPXV203 makes contacts to three distinct MHCI domains , pH regulation of the complex half-life is critically dependent on CPXV203 His residues that bind to an α3-domain acidic CD loop important for both Tpn and CD8 association . Thus , CPXV203 exploits a cellular pathway to target MHCI surfaces critical for immunological function in a manner that selects for those MHCI molecules most likely to present viral peptides . Remarkably , CPX203 is not related to any other MHCI-binding protein , but rather it is most structurally related to poxvirus CKBPs . To our knowledge , CPXV203 is the first member of the large T4 poxvirus protein family [38] to be structurally characterized , suggesting a previously underappreciated link between the poxvirus CKBPs and T4 protein families through similarities in their β-sandwich core . Regardless of the evolutionary history , the adaptation of this protein domain to structurally distinct ligands and unrelated functional outcomes suggests the integral role that CPXV203 plays in antigen presentation disruption may not be its only function . CPXV203 evolved into a promiscuous MHCI-binding protein by targeting MHCI determinants that are largely conserved by virtue of their roles in the recognition by host factors essential to cellular immunity ( Tpn , CD8 , NKRs ) . For instance , many α2-domain contacts ( R111 , Q115 , E128 , T134 ) are conserved through Tpn ( α2 128–136 ) [37] , [39] and Ly49 ( R111 , Q115 , D122 ) [40] interactions , while the β2m contacts primarily involve structurally conserved backbone positions within the LIR-1/MHCI interface [41] . The direct overlap of CD8 and Tpn contact sites ( Q226 , D227 , E229 ) [35]–[37] in the acidic CD loop of the α3 domain is clearly exploited by CPXV203 for MHCI binding , and this interface is precisely where we have identified two His residues in the viral protein that regulate increased kinetic stability at the lower pH of the Golgi . Previous investigations of pH-dependent endosomal ( PRL/PRLr [42] , FcRn/Fc [43] ) and ER→Golgi ( RAP/LRP ) [44] trafficking have repeatedly identified His residues as the pH sensitive component of these regulatory mechanisms . Unlike other amino acids , histidine is well suited to serve this function , as small pH shifts can drastically change the charge and hydrogen-bonding potential of this residue . As such , our investigation of CPXV203/MHCI pH regulation focused on interface histidines , which revealed a significant contribution of CPXV203 H75 and H80 to complex half-life at low pH . We suggest that these titratable His residues endow CPXV203 with the ability to regulate fully assembled MHCI in a manner that is complementary to the regulation of PLC-associated MHCI by CPXV012 . The specific binding of CPXV203 to fully assembled MHCI proteins in a pH-dependent manner clarifies mechanistically how CPXV203 coordinates with CPXV012 to effectively block antigen presentation . Previous characterizations showed the CPXV012 functions in a PLC-dependent fashion to block TAP transport of peptide into the ER [29] , [30] . However , some MHCI-binding peptides in the ER are not TAP-dependent and the CPXV012 block of peptide transport is likely not absolute . The MHCI proteins that are able to bind peptide in the presence of CPXV012 are left to CPXV203 , since it binds fully assembled MHCI through domain-specific conformational determinants conserved in classical and many non-classical MHCI . Among these interactions , the α3 interface is particularly important based on the presence of CPXV203 His residues that impart pH regulation to the CPXV203/MHCI interaction . This pH dependence suggests that CPXV203/MHCI interacts most avidly in the Golgi and not the ER , thus limiting the pool of MHCI that CPXV203 must retrieve . Interestingly , CRT has a C-terminal KDEL sequence conferring ER retrieval , and non-PLC-associated CRT has recently been implicated in quality control of MHCI peptide loading [14] . More specifically , CRT was shown to accumulate in the cis-Golgi and return peptide accessible MHCI proteins to the ER . Both CRT and CPXV203 retrieve MHCI proteins but with opposite goals . CRT functions in host quality control by retrieving MHCI with suboptimal peptides , whereas CPXV203 functions in immune evasion by retrieving fully assembled MHCI to block antigen presentation . Thus CPXV012 and -203 act sequentially to efficiently block MHCI expression using PLC-dependent versus PLC-independent mechanisms , respectively . As a possible consequence of efficient MHCI downregulation resulting in NK cell susceptibility , CPXV expresses the soluble class I-like protein OMCP that functions as a competitive antagonist of the NKG2D-activating receptor [24] . Indeed , the combined sabotage of both CTL and NK cell detection of virus-infected cells explains why mutant CPXV lacking CPXV012 and 203 demonstrates attenuated virulence in vivo compared to wild-type virus [29] .
MHCI-specific MAbs used in SPR competition assays were obtained from the ATCC ( H-2Kb: 25-D1 . 1 . 6 , B8-24-3 , Y-3 ) , purchased from BioLegend ( Kb: AF6-88 . 5 . 3 , Kb/Db: 28-8-6 , Kd/Dd: 34-1-2S , hβ2m: 2M2 ) , or provided as a kind gift ( Kb: 5F1-2-14 ) from S . Nathenson ( Albert Einstein College of Medicine , New York ) and L . Pease ( Mayo Clinic , Minnesota ) . MAbs that were not purchased from BioLegend were purified from ascites on a Bio-Rad Profinia FPLC using Protein A or G . MAbs used in flow-cytometry and IP assays have been described previously . MAb footprints in Figure S2H are based on SPR data from this work and available literature ( Text S1 ) . Peptides were synthesized by Fmoc chemistry and then subjected to reverse-phase HPLC for purification . Peptides were resuspended at >1 mM in ddH2O , DMSO , or 6M GuHCl , as dictated by peptide solubility . Peptides were chosen based on available MHCI crystal structures or personal suggestions by A . Stout ( NIH Tetramer Core Facility ) . See Text S1 for a list of all peptides used in this study . Murine embryo fibroblast ( MEF ) B6/WT3 ( WT3 ) and mutant MEFs including TAP1-deficient cells ( FT1− ) , Tapasin-deficient cells ( Tpn−/− ) , calreticulin-deficient cells ( CRT−/− ) , β2m-deficient cells ( B6 . B2M− ) and triple knockout fibroblasts ( Kb−/− Db−/− β2m−/−; 3KO ) were all derived from C57BL/6 ( H-2b ) embryos and have been described previously [45] . The CPXV203 and Kb mutants were stably expressed in the indicated cells by retroviral expression vectors pMXsIG [28] and pMIN [45] , respectively . Cells transduced by pMIN were selected by neomycin while green fluorescent protein ( GFP+ ) cells from pMXsIG transduced lines were enriched by cell sorting . For co-IPs ( TAP1/Tpn and CPXV203-HA/H-2Kb ) , cells were lysed in PBS with 1 . 0% digitonin ( Wako ) and protease inhibitor cocktails ( Roche ) for 60 min . Post-nuclear lysates were then incubated with indicated antibodies + protein A-sepharose ( Sigma ) or anti-HA sepharose ( sigma ) for HA-tagged CPXV203 for 1 h . After washes , coprecipitated proteins were eluted by boiling in lithium dodecyl sulfate ( LDS ) sample buffer ( Invitrogen ) . For cross-linking treatment ( CPXV203-HA/H-2Kb ) , cells were incubated with 1–2 mM DSP ( Thermo ) in PBS for 2 h at 4°C . The cross-linking was terminated with 25 mM Tris-HCl pH 7 . 4 before the cells were lysed in PBS with 1 . 0% NP-40 . Following immunoprecipitation cross-linked proteins were eluted by boiling in LDS sample butter with 2 . 5% β-mercaptoethanol . Immunoblot of precipitated proteins was performed following SDS-PAGE separation . Specific proteins were visualized by chemiluminescence using the ECL system ( Thermo ) . All flow cytometric analyses were performed using a FACS Calibur ( Becton Dickinson ) . Data were analyzed using FlowJo software ( Tree Star ) . Staining was performed as described [46] . PE-conjugated goat anti-mouse IgG ( BD Pharmingen ) was used to visualize MHCI staining . PE-conjugated anti-mouse CD1d ( eBioscience ) was used to detect surface CD1d . GFP signal representing CPXV203 transduced cells were collected in the FITC channel . Mammalian CPXV203ΔKTEL ( aa 1–205 , etgMVI-LHV ) , bacterial CPXV203ΔKTEL ( aa 1–205 , maMVI-LHV ) , and bacterial MHCI were produced using established methods ( Text S1 ) . Kb ( aa 0–280 , mGPH-PST ) /H-2Kd ( aa 0–283 , mGPH-VSN ) constructs were produced in house , while H-2Dd , H-2Dk , H-2Dq , H-2Kk , H-2Ld , Mamu-A*01 , Patr-B*0802 , H-2Q9-H-2Db , H-2Kb-HLA-A*0201 , HLA-A*0201-H-2Kb were produced by the NIH Tetramer Core Facility . H-2Db biotinylated monomer ( LCMV Gp33 , KAVYNFATC ) was purchased from Beckman Coulter . Biotinylated constructs included a C-terminal site-specific biotinylation tag and were biotinylated by following established procedures ( Avidity ) . Chimeric mCD1d-Fc produced in a murine cell line was obtained from R&D Systems . All MHCI include human β2m ( hβ2m , aa 0–99 , mIQR-RDM , UNP: P61769 ) unless otherwise noted . Signal-peptide/cloning artifacts are indicated as lower-case aa . CPXV203 ( Brighton Red strain , SeMet-labeled ) /MHCI ( OVA257–264:H-2Kb:hβ2m ) complex was prepared for crystallization by size-exclusion chromatography purification of CPXV203/MHCI in low pH/salt buffer ( 50 mM NaCl , 30 mM MES pH 5 . 6 , 0 . 01% Azide ) . Diffraction-quality crystals of CPXV203/MHCI were grown at 20°C by streak seeding into hanging drops of 0 . 5 µl complex ( 7 . 5 mg/ml ) +0 . 5 µl reservoir solution ( 10% PEG 6000 , 4% glucose , 2% ethylene glycol , 0 . 1 M tri-K citrate pH 5 . 55 , 0 . 01% Azide ) . Crystals were dehydrated , flash frozen in liquid N2 , and then used for X-ray data collection at the Advanced Light Source ( ALS ) beamline 4 . 2 . 2 ( 0 . 97909 Å wavelength ) . Crystals belong to space group P1 ( a = 88 . 31 Å , b = 88 . 25 Å , c = 106 . 42 Å; α = 76 . 18° , β = 69 . 29° , γ = 66 . 69° ) with four CPXV203/MHCI complexes per asymmetric unit ( ASU ) and a solvent content of 52% . The HKL2000 software package [47] was used to index , integrate , and scale the data , yielding an 87 . 9% complete dataset at 3 . 0 Å ( R-sym = 16 . 4% ) ( Table S2 ) . The structure of H-2Kbm8:mβ2m ( 2CLZ ) [48] was used for molecular replacement ( MR ) using Phaser within the Phenix suite [49] , with four MHCI proteins located within the ASU . Electron density for the unique CPXV203 region was improved by cross-crystal averaging [33] allowing an initial model of CPXV203 to be manually built in Coot [50] . MR-SAD using AutoSol ( Phenix suite [49] ) was subsequently used to identify eight SeMet sites in the four CPXV203 monomers , enabling the introduction of anomalous phase information ( figure of merit [FOM] 0 . 604 ) that improved map quality allowing for complete CPXV203/MHCI model building . Phenix Refine [49] was used with global non-crystallographic symmetry ( NCS ) restraints to refine the CPXV203/MHCI structure to a final Rwork of 22 . 9% and Rfree 25 . 3% ( see Table S2 for complete crystallographic statistics ) . The final CPXV203/MHCI model contains four CPXV203/MHCI complexes and 72 water molecules . Each complex contains the nearly complete CPXV203 ( mature residues 5–190 ) , Kb ( mature residue 1–277 ) , human β2m ( 0M-99 ) , and OVA257–264 . All structure figures were created using PyMOL [51] . See Text S1 for comprehensive crystallization , data collection , structure determination , refinement , structural analysis , and figure preparation details . SPR experiments were run on a Biacore T100 ( GE Healthcare ) in either standard HBS-EP+ ( pH 7 . 4 ) or low pH MBS-EP+ ( pH 6 . 0 ) . Kinetic and equilibrium analyses were performed using Biacore T100 Evaluation software using a 1∶1 Langmuir model . BLI experiments were performed on an Octet RED system ( ForteBio ) using HBS-EP+/MBS-EP+ supplemented with 0 . 05% ( v/v ) TWEEN and 1% BSA . Equilibrium BLI data was analyzed using Octet software ( V7 . 0 ) . All biosensor experiments were run at 25°C and followed proper biosensor experimental technique . Size-exclusion chromatography-multi-angle light scattering ( SEC-MALS ) experiments were run on a Dawn HELEOS-II 18-angle light scattering detector ( Wyatt ) and Optilab rEX refractive index monitor ( Wyatt ) linked to a Waters HPLC system . Dynamic light scattering ( DLS ) was performed on a DynaPro-801TC . Circular dichroism was measured using a Jasco-810 instrument ( Jasco Inc . ) . Detailed methodologies for biosensor , light scattering , and circular dichroism experiments are available in Text S1 . | Viruses encode a wide array of proteins whose principle function is to disable the surveillance and effector functions of the immune system . A common viral target is the MHC class I antigen processing and presentation pathway , which is a potent mechanism used by the host for the detection and killing of virally infected cells . In this study we have delineated the immune evasion mechanism of the cowpox-encoded CPXV203 protein , which is known to potently block the normal trafficking of MHCI from the endoplasmic reticulum ( ER ) to the plasma membrane . CPXV203 does this by highjacking an ER-retrieval system that usually serves to capture defective , chaperone complexed MHCI proteins in the Golgi and send them to the ER . We show that CPXV203 adopts a compact beta-sandwich structure and engages evolutionarily conserved MHCI determinants that are located under the peptide-binding platform . The viral protein binds a variety of different MHCI proteins weakly at the pH found in the ER , but the affinity and half-life are significantly augmented at the more acidic conditions found in the Golgi . Together these data suggest that CPXV203 works cooperatively with the endogenous ER retrieval process to promiscuously target fully assembled MHCI , thereby preventing T-cell killing of cowpox infected cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biomacromolecule-ligand",
"interactions",
"major",
"histocompatibility",
"complex",
"antigen",
"processing",
"and",
"recognition",
"viral",
"immune",
"evasion",
"virology",
"immunology",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"biophysics"
] | 2012 | Structural Mechanism of ER Retrieval of MHC Class I by Cowpox |
We have identified from the mutualistic grass endophyte Epichloë festucae a non-ribosomal peptide synthetase gene ( sidN ) encoding a siderophore synthetase . The enzymatic product of SidN is shown to be a novel extracellular siderophore designated as epichloënin A , related to ferrirubin from the ferrichrome family . Targeted gene disruption of sidN eliminated biosynthesis of epichloënin A in vitro and in planta . During iron-depleted axenic growth , ΔsidN mutants accumulated the pathway intermediate N5-trans-anhydromevalonyl-N5-hydroxyornithine ( trans-AMHO ) , displayed sensitivity to oxidative stress and showed deficiencies in both polarized hyphal growth and sporulation . Infection of Lolium perenne ( perennial ryegrass ) with ΔsidN mutants resulted in perturbations of the endophyte-grass symbioses . Deviations from the characteristic tightly regulated synchronous growth of the fungus with its plant partner were observed and infected plants were stunted . Analysis of these plants by light and transmission electron microscopy revealed abnormalities in the distribution and localization of ΔsidN mutant hyphae as well as deformities in hyphal ultrastructure . We hypothesize that lack of epichloënin A alters iron homeostasis of the symbiotum , changing it from mutually beneficial to antagonistic . Iron itself or epichloënin A may serve as an important molecular/cellular signal for controlling fungal growth and hence the symbiotic interaction .
Iron is an essential nutrient for almost all organisms ( except for some lactobacilli [1] ) due to its central role in vital cellular reactions . The redox properties of iron confer a catalytic function essential for fundamental metabolic pathways such as DNA synthesis , respiration and photosynthesis [2] , [3] . Even though iron is a highly abundant metal , in aerobic environments bioavailability is low because ferric iron forms insoluble oxides through reaction with oxygen . Iron concentrations in tissues must also be carefully regulated since free iron is deleterious given that it has the potential to catalyze the production of cytotoxic reactive oxygen species ( ROS ) through the Haber-Weiss/Fenton reactions [4] . Controlling iron homeostasis is therefore an essential function and organisms have developed complex strategies for iron uptake , utilization , transport and storage [5] . The uptake of iron is presumably the principal regulatory point of iron homeostasis and multiple high- and low-affinity iron uptake systems have evolved [6] . In mammals and plants excess iron is tightly sequestered by high-affinity binding proteins [7] , [8] , and deliberately withdrawing iron is a documented defense strategy employed by animal hosts against invading bacterial pathogens in order to limit their growth [9] , [10] . For microbes to acquire limited iron from animal or plant hosts , high affinity iron uptake systems are required . One such mechanism is siderophore-mediated iron uptake . Under iron starvation , most fungi and bacteria synthesize siderophores , low molecular weight ( Mr<1 , 500 ) iron-chelating agents , to solubilize ferric iron and control their intracellular iron levels [11]–[13] . The functional role of siderophores may be extracellular – secreted as iron-free siderophores to chelate iron for cellular iron uptake; and/or intracellular – located within the cell for intracellular iron storage . A number of roles have been attributed to fungal siderophores and these include functions such as virulence factors and asexual/sexual determinants [14]–[21] . Another major type of high affinity iron acquisition in fungi is reductive iron assimilation ( RIA ) which is based on a ferroxidation/permeation uptake [6] , [22] , [23] . Some fungi are capable of utilizing both of these high affinity uptake systems; examples are Ustilago maydis , Aspergillus fumigatus and Fusarium graminearum [15] , [16] , [24]–[28] . Nevertheless , characterization studies of both siderophore and RIA uptake systems in the same fungal species through respective gene deletion studies have shown that only one of these two systems is essential for mammalian or plant disease by a given fungal pathogen and which one is indispensable thus appears to be fungal species dependent and possibly related to their lifestyles . Examples of fungi demonstrated to require the siderophore system for virulence are A . fumigatus , Histoplasma capsulatum and F . graminearum [16] , [19] , [29] , whereas RIA is essential for the virulence of the basidomycetes U . maydis , Candida albicans and Cryptococcus neoformans [15] , [30] , [31] . Nearly all fungi produce hydroxamate-type siderophores [32] , which are typically composed of three hydroxamate groups linked by peptide or ester bonds to form an octahedral complex . Siderophores are classified into four major structural types: ferrichromes , fusarinines , coprogens and rhodotorulic acid [12] . Their formation by the linkage of hydroxamate groups or additional amino acids in the case of ferrichrome-type siderophores is catalyzed by non-ribosomal peptide synthetases ( NRPS ) [28] . NRPSs are multifunctional enzymes that synthesize peptides by a thio-template mechanism and are modular in structure . A typical module consists of an adenylation ( A ) domain for substrate specificity , a peptidyl carrier domain ( T ) , which binds a 4′-phosphopantetheine cofactor , and a condensation ( C ) domain for bond formation [33] . NRPS genes functionally identified to encode fungal siderophore synthetases with an extracellular role are sid2 and fer3 of U . maydis and sib1 of Schizosacccharomyces pombe , involved in the synthesis of ferrichrome-type siderophores [28] , [34] , [35] , sidD from A . fumigatus involved in triacetylfusarinine C ( TAFC ) production [36] as well as NPS6 of Alternaria brassicicola , F . graminearum , and Cochliobolus heterostrophus involved in the respective production of Nα-dimethylcoprogen , TAFC and coprogens [19] . NPS6 deletion studies have demonstrated that these extracellular siderophores play a role in fungal virulence to plants [19] . Recently , another NPS6 ortholog responsible for the production of coprogens was identified from the plant pathogenic fungus Magnaporthe grisea , but loss of the corresponding gene did not affect virulence in rice [37] , whereas the gene encoding the NRPS responsible for the synthesis of the intracellular siderophore , ferricrocin , was required for full virulence of M . grisea [20] . Fungi that form mutualistic relationships with plants such as the widespread symbioses of mycorrhizal fungi with terrestrial plant communities also produce siderophores [38]–[40] . In contrast to fungal pathogens , siderophore production by mycorrhizal fungi has been postulated to aid their hosts by improving solubilization of insoluble iron oxides resulting in a positive effect on the iron nutrition of the host plant [41] . The epichloae fungi of family Clavicipitaceae ( comprising genera Epichloë and Neotyphodium ) form symbioses with temperate grasses of the subfamily Poöideae [42] . These Poöideae-epichloae associations comprise an evolutionary continuum from mutualistic to antagonistic , with the nature of this relationship determined by the importance of vertical ( via host seeds ) versus horizontal ( ascospore mediated ) transmission of the fungus [42] . During colonisation of grass leaves , the endosymbiont's growth is tightly regulated and synchronized with the growth of its host plant [43] . Fungal hyphae are confined to the intercellular spaces of leaf sheaths and blades where the endophyte induces no symptoms [44] . Epichloid symbioses can be mutually beneficial with the plant providing the endophyte with nutrients , and the endophyte conferring biotic and abiotic advantages to the host plant [44]–[48] . Improved herbivore resistance of infected plants is linked to the production of fungal secondary metabolites ( alkaloids ) , and it appears that the host plant plays a key role in the regulation of some of these metabolites [49] , [50] . In a study to investigate the distribution and diversity of NRPS gene families within the epichloae , two NRPS fragments , NRPS2 and NRPS9 , were isolated with proposed functions identified as siderophore-like [51] . Our unpublished data shows that NRPS9 encodes sidC , the siderophore synthetase for ferricrocin ( L . Johnson et al . , unpublished results ) . NRPS2 , renamed as sidN , was confirmed to be a siderophore-synthesizing NRPS from three-dimensional structural elucidation and biochemical studies of the third adenylation domain of SidN [52] . These studies showed that this domain activates anhydromevalonyl-N6-hydroxy-L-ornithine ( AMHO ) , a large hydroxamate amino acid known to be a component of fungal siderophores [52] . More recently , the structure of a novel ferrichrome-type siderophore containing AMHO moieties which we have designated as epichloenin A , has been elucidated by high resolution tandem mass spectrometry ( HRMSMS ) and NMR from iron-depleted cultures of E . festucae [53] . An additional minor variant , epichloënin B , has also been detected in cultures of WT and C-sidN strains [53] . Here , we report a study establishing that sidN ( formerly NRPS2 ) encodes a siderophore synthetase required for the biosynthesis of the extracellular siderophore , epichloënin A . Our research shows that sidN is required for maintaining the mutualistic interaction of the endophyte E . festucae with perennial ryegrass .
To investigate whether sidN ( formally NRPS2 ) is involved in the biosynthesis of epichloënin A and its role in endophyte-grass symbioses , a gene disruption construct was designed and the gene disruption performed by homologous recombination in E . festucae wild-type ( WT ) strain Fl1 ( see Methods for details ) . Four independent ΔsidN mutants ( ΔsidN 85 , 82 , 54 , and 26 ) were identified at a homologous recombination frequency of 10% . These four mutants displayed similar phenotypes as reported throughout this study . The axenic vegetative growth characteristics of the ΔsidN E . festucae mutants were examined on defined medium ( DM ) and on DM supplemented with the ferrous iron chelator bathophenanthrolinedisulfonic acid ( BPS ) , which is impermeable to cell membranes [54] , [55] ( Figure 1A ) . BPS supplementation sequesters trace ferrous iron from the extracellular medium , subsequently inhibiting RIA iron uptake by removing the substrate for this process . For the WT , addition of BPS to DM moderately reduced the radial growth rate ( by approximately 20% ) , whereas radial growth of the ΔsidN mutants was almost completely inhibited ( shown for two independent ΔsidN mutants in Figure 1A ) . The lack of growth by the mutants on this medium is explainable by the loss of both RIA-mediated ( via ferrous iron BPS chelation ) and siderophore mediated iron uptake from the extracellular medium ( i . e . loss of biosynthesis of the extracellular siderophore ) , and demonstrates the presence of RIA in E . festucae . The inhibition of radial growth on DM+BPS medium could be moderately complemented by the addition of enriched culture supernatants ( Sephadex-column purified culture filtrate - CF ) from WT ( Figure 1A ) , but not from ΔsidN 85 ( data not shown ) , indicating the presence of the native extracellular siderophore in WT CF only . Growth of all ΔsidN mutant strains could also be fully restored by addition of FeSO4 , or FeCl3 to DM+BPS medium ( Figure 1A ) . To validate that the phenotypic effects described in the ΔsidN strains arose from gene inactivation of sidN , a complementation was performed by co-transforming protoplasts from ΔsidN 85 with fosmid clone G8-5 ( containing the full-length genomic copy of E . festucae sidN ) and a plasmid containing the geneticin resistance gene ( pII99 ) . Screening for complementation was carried out using an in vitro plate assay that exploited the phenotypic trait of the ΔsidN mutants to have extremely poor mycelial growth on BPS-containing , iron-depleted DM ( see Figure 1A ) . Four independent transformants formed large colonies on BPS medium . We designated the complemented strain used for further studies C-sidN; measurements of radial growth of ΔsidN 85 , WT and C-sidN show that C-sidN is able to grow at 94% of WT on BPS supplemented DM ( Figure 1B ) . This indicates that transformation with the full-length genomic copy of sidN restored growth of ΔsidN mutants on BPS supplemented medium and congruent with sidN encoding a biosynthetic enzyme for high-affinity siderophore uptake of iron from extremely iron-depleted medium . To determine if the siderophore assembled by the biosynthetic pathway involving SidN was epichloënin A , a chemical analysis was performed on cultures of the WT , ΔsidN mutant and complemented C-sidN strains grown under iron depleted conditions ( Figure 2 ) . Culture supernatant and mycelial extracts from these strains were analyzed by liquid chromatography – mass spectrometry ( LCMS ) . Epichloënin A ( MS1 m/z 542 ( [MH2]2+ ) was the predominant form in the supernatant of WT cultures ( Figure 2 ) , while ferriepichloënin A ( MS1 m/z 569 ( [MH2]2+ ) predominated in the mycelial extracts . Since the ratio of the iron-free to iron-bound form is much higher in the extracellular supernatant than in the mycelium , we conclude that epichloënin A is an extracellular siderophore . A similar pattern of occurrence was observed for the C-sidN supernatants , albeit at reduced concentrations compared to WT . However , both forms were absent from supernatants and mycelia of the ΔsidN mutant strains , clearly demonstrating that SidN is responsible for the assembly of epichloënin A . Figure 3A presents the postulated siderophore biosynthetic pathway according to Plattner and Diekmann [56] adapted for epichloënin A biosynthesis in E . festucae . To enable further investigation of sidN ( formally NRPS2 ) , including the determination of the modular architecture of this siderophore-synthesizing NRPS , we obtained the full-length gene sequence by constructing an E . festucae ( strain Fl1 ) fosmid library . Screening of this library by PCR ( using NRPS2 specific primers ) gave three positive clones , of which one ( clone G8-5 ) contained the entire genomic copy of the full-length gene . Sequencing of the E . festucae Fl1 sidN fosmid clone revealed an open reading frame ( ORF ) of 14 , 073 bp with no introns . A BLASTX analysis of the deduced amino acid sequence of sidN revealed similarities to fungal NRPS genes involved in siderophore biosynthesis . Top hits to SidN ( with an expect value of 0 . 0 ) included functionally characterized siderophore NRPSs involved in the biosynthesis of ferricrocin or ferrichrome . The top hit was to A . fumigatus SidC [36] at 34% amino acid identity ( 52% positives ) . Out of the six modular architectures described for ferrichrome synthetase NRPSs by Bushley et al . [57] , the most significant hits using sidN as the query sequence were to 12 of the 13 Type II ferrichrome NRPSs , followed by less similarity to other ferrichromes . These results support our previous findings that SidN encodes a siderophore biosynthetic enzyme of the ferrichrome family [52] , and upholds our current discovery that SidN assembles epichloënin A , a novel ferrichrome family member . It is however interesting that the amino acid sequence analyses indicates SidN has the highest similarity to SidC which assembles ferricrocin , but the acyl groups of the hydroxamates of both epichloënin A and ferricrocin are different: for SidN it is trans- anhydromevalonyl [52] , [53] , whereas for ferricrocin , a simple acetyl group is incorporated into its hydroxamate group [11] . Analysis of the SidN amino acid sequence revealed three complete A-T-C modules and a terminal T-C repeat identical to Type II ferrichrome NRPSs ( where A = adenylation domain , T = peptidyl carrier domain and C = condensation domain ) ( Figure 3B ) . From the work by Lee et al . [52] the third domain ( A3 ) of SidN incorporates the hydroxamate groups of the siderophore which forms an octahedral iron complex . The other component amino acids of epichloënin A are one glutamine and four glycines [53] , which we postulate are assembled by SidN A domains one ( A1 ) and two ( A2 ) . Analysis of the putative binding pocket residues of the SidN A1 and A2 domains showed that they were dissimilar to those of other characterized fungal siderophore synthetases [28] , [57] and did not allow a prediction of which A domain activated glutamine or glycine or the number of iterative cycles of their addition into the peptide . Figure 3B shows that A1 and A2 domains of SidN activate either the addition of one glutamine or four cycles of glycine addition , while the A3 domain is predicted to activate three cycles of addition of trans-AMHO residues . Recently , the genomes of two E . festucae strains ( Fl1 and E2368 ) have been sequenced ( http://www . endophyte . uky . edu/ ) . A BLASTN analysis of the E2368 and Fl1 sidN sequences revealed a 99% nucleotide identity with an expect value of 0 . 0 . We also previously identified a second siderophore-like NRPS from N . lolii ( NRPS9; [51] ) which was located on a different supercontig to sidN in both E . festucae genomes and putatively encodes an intracellular siderophore synthetase , ferricrocin ( L . Johnson et al . , unpublished results ) . We have not identified any other siderophore-like NRPS genes in the E . festucae Fl1 and E2368 genomes . To determine what other genes are clustered with sidN we analyzed the predicted ORFs surrounding the sidN homologue from the E . festucae Fl1 and E2368 genome databases . These analyses revealed that just one ORF ( abcI ) , located upstream of sidN appears to be grouped with sidN forming a partial siderophore biosynthetic gene cluster , possibly sharing a divergent promoter with sidN . This ORF encodes a putative member belonging to the P-glycoprotein-like multidrug resistance ( MDR ) subfamily of ATP-binding cassette ( ABC ) transporters . A similarity search revealed a highly significant hit to ABC1 from H . capsulatum , located in the H . capsulatum SID1 gene cluster in an analogous position and orientation with respect to its siderophore synthetase , NPS1 [29] . There were no other shared regions of synteny . Of note , ABC1 is iron-regulated and contains a putative Sre1 GATA motif consensus site , for regulation by the GATA type iron regulator [29] . Schrettl et al . [36] reported the accumulation of the siderophore precursor cis-AMHO , in the supernatant of an iron-depleted culture of a sidD deletion mutant; sidD encodes the NRPS responsible for fusarinine C biosynthesis in A . fumigatus . We therefore examined supernatants of cultures of ΔsidN mutants grown under iron-depleted conditions by LCMSMS for the presence of likely precursors of epichloënin A , and found a product of m/z 261 which we identified putatively as trans-AMHO by comparison with the retention and MS2 spectrum of cis-AMHO prepared from fusarinine C [52] ( Figure S1A ) . This is consistent with the finding that epichloënin A incorporates trans-AMHO moieties [53] . Targeted LCMSMS analyses of mycelial extracts of the WT , ΔsidN 85 mutant and the complemented strain showed that trans-AMHO accumulation was very low in the WT , slightly higher in the complemented strain , but nearly 40-fold higher in ΔsidN 85 compared to WT , indicating that trans-AMHO is a precursor that is significantly accumulating in the mutant ( Figure S1B ) . Light microscopy was employed to observe the vegetative hyphal growth characteristics of the ΔsidN mutants on water agar as it is a useful medium for studying the morphology of individual hyphal strands within a colony . It is also a nutrient deficient medium . The mutants displayed various abnormalities compared to WT such as increased lateral branching , coils ( not shown ) , atypical hyphal convolutions with a high frequency of hyphal tip and compartmental swellings ( Figure 4 ) ; complementation in the C-sidN strain restored normal hyphal growth ( data not shown ) . To evaluate the cell wall integrity of the ΔsidN mutants , staining with Calcofluor White was used to analyze the distribution of chitin ( and other β-linked fungal cell wall polysaccharides ) [58] in hyphae grown in liquid DM , and DM amended with BPS or 20 µM iron ( Figure 5 ) . For the WT , there were no significant differences in tip , septa , or hyphal wall staining in DM or DM supplemented with iron , however , the addition of the ferrous iron chelator BPS to the medium resulted in the majority of tips displaying no or weak speckled staining near the tip region , with septa staining not affected . Also , the hyphal contents appeared to have a much higher background stain compared to staining of hyphae in DM or DM with iron supplementation . The complement was comparable to the WT ( results not shown ) . It is remarkable that addition of BPS , an agent that chelates ferrous iron and as a consequence inhibits RIA uptake , can dramatically alter chitin localization . A similar response to BPS supplementation was observed with stained mycelium from ΔsidN 85 , yet some specific differences with respect to WT were observed under all 3 conditions tested . In DM , mutant hyphae displayed slightly less tip staining compared to the WT , and often the length of staining at the tip was reduced ( Figure 5 ) . Addition of iron appeared to improve overall tip staining length and brightness to that observed for WT hyphae . Supplementation with BPS , also radically affected the chitin distribution of mutant hyphae ( similar to WT ) , however the pattern of hyphal growth on this medium was highly branched ( as observed for ΔsidN mutants on water agar ) , and notably some of these branches were extending in a meandering manner ( see Figure 5B ) . Collectively , these observations suggest that the ΔsidN mutants are defective in maintaining polar growth when iron availability is low; to our knowledge , this is a feature not previously linked to fungal siderophore mutants . To test for sensitivity of the ΔsidN mutants to oxidative stress , hydrogen peroxide was applied to the growth medium . On complete medium ( PD ) , only ΔsidN mutant 85 showed a slight sensitivity to H2O2 , whereas on iron-depleted DM both ΔsidN mutants 54 and 85 were significantly more sensitive to H2O2 than WT ( Table 1 ) . A reduction in asexual sporulation ( in iron-depleted medium ) had been previously documented for deletion strains of C . heterostrophus NPS6 [18] . To study sporulation in the ΔsidN mutants , it was necessary to find suitable asexual sporulation conditions to generate sufficient numbers of conidia from the WT strain for comparison . Only growth on water-agar ( with a cold-inducing incubation step ) produced satisfactory levels of sporulation in the WT on coil structures and hyphal strands , while in comparison , colonies of the ΔsidN mutants produced very low numbers of conidia under these conditions ( see Figure S2A and S2B ) . Asexual spore morphology did not appear to be affected by the sidN mutation ( Figure S2B ) . It is not currently possible to determine if ascospore formation is affected in these mutants since the pre-sexual choke state required for the commencement of the sexual cycle has never been observed with artificial infection of perennial ryegrass ( L . perenne ) with E . festucae [59] . Fungal siderophore biosynthetic genes are typically negatively regulated by iron , and we therefore compared the expression of sidN from liquid cultures grown under iron depleted conditions versus iron supplemented conditions . Expression of sidN in the WT was strongly repressed under iron supplemented conditions and only detectable by reverse transcriptase polymerase chain reaction ( RT-PCR ) when grown in iron-depleted medium ( Figure 6 ) . As expected , expression of sidN was not detected in the ΔsidN mutants under any of the conditions tested . To explore the effect of loss of the sidN product on the symbiotic relationship of E . festucae with perennial ryegrass , ΔsidN mutants , WT and C-sidN complemented strains were inoculated into perennial ryegrass seedling lines ( Figure 7A ) . Onset of symptom development with the ΔsidN mutants was apparent within two weeks of planting inoculated seedlings into soil . Infection with the ΔsidN mutants resulted in a range of stunted , erect ( versus prostrate ) plant phenotypes , all with a large increase in tiller numbers ( Figure 7A ) . All plants infected with ΔsidN mutants showed abnormal underdeveloped root systems , and with the more severely stunted plants having markedly reduced root systems ( Figure 7B ) . Severely stunted ΔsidN infected plants frequently died when transplanted from root trainers to larger pots , in contrast to the more moderately stunted plants , which could be maintained for long periods in the glasshouse . To examine hyphal growth in planta , light microscopy of aniline-blue stained leaf sheaths of grass pseudostems was performed ( Figure 7C , D , E ) . Hyphae of WT infected plants grow mainly by intercalary growth typified by infrequently branched hyphae that are generally aligned parallel to the longitudinal leaf axis of the elongating grass leaf [43] . Microscopic examination of ΔsidN mutants in perennial ryegrass revealed extensive colonization of leaf sheath tissues in comparison to WT ( Figure 7C , D , E ) . There were large regional areas of synchronized hyphal growth ( as described by Christensen et al . [43] ) alongside unrestricted convoluted hyphae ( Figure 7D , E ) . Less frequent patches of excessively convoluted hyphae extending out in all directions were also observed ( see Figure 7E for an example ) . Toluidine blue stained 1 µM cross sections of pseuodostems from WT , C-sidN and ΔsidN mutants 54 and 85 were examined to analyze the distribution of hyphae within different ages of leaf sheaths and blades ( Figure 8A , B ) . We observed that hyphae from the WT or the complemented strains showed an even distribution of hyphae across all tissue types and tissue ages , but that hyphae from stunted plants infected with ΔsidN 54 or 85 were abnormally distributed . Specifically , the number of ΔsidN mutant hyphae were found to be the lowest in the inner developing leaf blade , and increased with age so that more hyphae were observed in the older outer leaf blade and sheaths ( results not shown ) . We could also gain information about the cytoplasmic density of hyphae in specific tissues based on penetration of toluidine blue . Hyphae from the WT or complemented strain were generally dense in all tissues examined , but not so for the mutant strains . Dense hyphae were only observed in ΔsidN 54 or 85 infections from within the youngest tissue , the inner leaf blade , as well as hyphae located in the phloem of the vascular bundles . However , the majority of mutant hyphae located in the older leaf blade and all three sheaths were empty ( Figure 8A shows hyphae located in the inner leaf sheath in both mesophyll and vascular tissue ) . Unlike the WT in which no hyphae were observed in the vasculature , mutant hyphae were found in large numbers in some of the vascular bundles ( mostly in the phloem ) of all of the tissue types examined ( Figure 8A ) . Additionally , we noted many epiphyllous hyphae present with the mutant infection compared to few for the WT or complement , with most being large in diameter and apparently empty as hyphal contents were not stained at all with toluidine blue ( Figure 8A ) . These light microscopy pictures indicate that the ΔsidN mutant is proliferating mostly in only the older parts of the plant , with typical WT-like growth confined to the inner leaf blade , where hyphae are few and stain densely . Generally , the densely-stained mutant hyphae were only seen in nutrient rich vasculature or in the inner leaf blade . Transmission electron microscopy ( TEM ) of perennial ryegrass plants infected with WT , complement and ΔsidN mutants 54 and 85 confirmed the light microscopy findings ( described above ) and provided detailed images of various abnormalities apparent in the hyphal ultrastructure of ΔsidN mutants ( Figure 8B ) . Frequently observed irregularities were atypical vacuolation as well as poorly stained hyphae versus rarely vacuolated and frequently densely-stained WT/complemented hyphae indicating that a characteristic feature of the hyphae of ΔsidN mutants is diffuse cytoplasm ( Figure 8B , Figure 9 ) . As already noted by light microscopy , hyphae located in the vasculature contained dense cytoplasm , and the contents as determined by TEM appeared very normal ( WT-like ) and hyphae were also mostly regularly shaped like WT ( Figure 9 ) . Additionally , we often observed asymmetrically shaped mutant hyphae of variable diameter in the mesophyll of various plant tissue types and occasionally , but only in mutant infected plants , the presence of clusters of hyphae orientated both longitudinally , obliquely and transversely surrounding host cells , but did not enter host cells ( Figure 9 ) . Epiphyllous hyphae were also able to be visualized by TEM in ΔsidN infections as they were so abundant , and were observed to always be highly vacuolated and encapsulated in plant host derived material ( Figure 9 ) . Light microscopy and TEM observations showed that the ΔsidN mutants colonized the apoplastic space with multiple hyphae compared to single or few hyphae in the WT ( Figures 8A , 8B , 9 ) . This apparent increase in fungal biomass was quantified by real-time quantitative PCR ( qPCR ) of the single copy fungal gene , NRPS1 , in total gDNA ( plant and fungal; [60] ) . Relative to WT infected plant pseudostems ( that comprise a mixture of enclosed immature emerging leaf blades and surrounding mature sheaths ) , the fungal biomass in plants infected with ΔsidN 54 was 1 . 4 fold higher , while in plants infected with ΔsidN 85 fungal biomass was 1 . 8 fold higher . To determine whether the fungal extracellular siderophore migrated into the plant , we analyzed guttation fluid from perennial ryegrass infected with WT , ΔsidN 85 mutant and complemented C-sidN strains by targeted LCMSn , as we have found that this fluid provides a clean matrix for the detection of endophyte metabolites in planta [61] . While epichloënin A could not be detected in planta , ferriepichloënin A was detectable at trace levels in several samples from plants infected with the WT or C-sidN strains but could not be detected in plants infected with the ΔsidN 85 mutant ( results not shown ) . Indirect effects of the ΔsidN 85 mutant on the chemical phenotype of the symbiotum were more marked . E . festucae strain Fl1 , in association with perennial ryegrass , synthesizes three major classes of alkaloids: indolediterpenoids , of which lolitrem B is the major product; ergot alkaloids , of which ergovaline is the major product , and the pyrrolopyrazine peramine . Production of all three is minimal or undetectable in axenic cultures , but all are produced in significant amounts ( of the order of µg/g ) in planta [62]–[64] . Many biotic and abiotic factors influence the production of fungal alkaloids in planta , including mineral stress , for example nitrogen and phosphorous availability [60] , [65] , [66] . To test if the altered phenotype of ΔsidN 85 infected plants may have an effect on alkaloid production , we determined alkaloid concentrations in infected plants on at least 3 occasions . While considerable variation was observed in concentrations of lolitrem B , and to a lesser extent of peramine , ergovaline consistently accumulated to much higher levels in the ΔsidN 85 mutant plants compared to WT ( 10-fold for ΔsidN 85 , Figure 10 ) , a factor much larger than the 1 . 8-fold increase in fungal biomass in the identical plant tissue as described above . Other ΔsidN mutant strains showed a similar elevation in ergovaline levels ( data not shown ) . In agreement with this , a transcriptomic experiment that compared plants either infected with WT or ΔsidN 54 strains by a custom designed Affymetrix GeneChip revealed that the expression of genes involved in the biosynthesis of ergovaline were highly elevated in the plants infected with the ΔsidN mutant ( L . Johnson et al . , unpublished results ) . Siderophore mediated iron-uptake does not appear to be the only high affinity iron uptake system operational in Epichloë as a search of the E . festucae ( E2368 , Fl1 ) genome sequences revealed the presence of a putative ferroxidase , fetC and a putative high affinity iron permease , ftrA that form the bipartite Fet3/Ftr1 complex responsible for reductive iron assimilation ( ferroxidation/permeation ) [23] in these endophyte strains . Only one putative ortholog of fetC and ftrA , respectively , was identified . These two genes are co-localized in the genome and share a bidirectional promoter as is commonly found for other fungi [15] , [16] , [67] . Two other genes regulating iron homeostasis were also found in the genome; sreA , the GATA-type transcriptional repressor of iron uptake during iron-replete conditions [68] , [69] , and hapX involved in the repression of iron-utilizing proteins under iron deficiency mediated via the CCAAT-binding complex [70]–[73] . See Table S2 for detailed descriptions of identified genes used for RT-qPCR below . To test our hypothesis that the ΔsidN mutants in perennial ryegrass , due to loss of their extracellular siderophore , are sensing changes in their cellular iron status , we quantified mRNA abundances of the putative E . festucae iron-repressed genes fet3 , ftr1 , and hapX . Significant up-regulation of all three iron-repressed genes in pseudostem tissue of ΔsidN 54 and 85 compared to WT infected plants was found , suggesting reduced intracellular iron levels in the ΔsidN mutants as a result of the loss of epichloënin A production ( Figure 11 ) . The roles of genes of the superoxide-generating NADPH oxidase ( Nox ) family have been well characterised in E . festucae and shown to be critical for symbiosis maintenance [74]–[77] . Perturbations in levels of reactive oxygen species ( ROS ) via mutation of components of the Nox complex in E . festucae affect regulation of hyphal growth and branching in the host and cause plant stunting . A number of shared phenotypic features of the in planta phenotypes of the ΔsidN mutants with the E . festucae Nox mutants , such as unrestricted hyphal growth and swollen hyphal tips , suggested that alterations in ROS could play a role in the ΔsidN mutant phenotype . To investigate whether there were any alterations in the transcriptional regulation of the Nox complex , we quantified mRNA abundances by RT-qPCR of the Nox genes , noxA ( required for symbiosis , [75] ) and noxB ( no known function , symbiosis not affected , [75] ) , as well as the regulators of this complex noxR [74] and the small GTPase racA ( required for NoxA activation and regulated ROS production , [76] ) in pseudostem tissue of plants either infected with WT or ΔsidN mutant strains . For perennial ryegrass plants infected with ΔsidN mutants , noxA , noxB and noxR genes were not differentially expressed relative to WT infected plants , but racA was down-regulated 2 . 2 to 3 . 3 fold in ΔsidN 85 and ΔsidN 54 infected plants respectively ( Figure 11 ) . To determine if altered racA expression resulted in a measureable difference in ROS production in the ΔsidN mutants , we examined ROS production in axenic cultures by both nitroblue tetrazolium ( NTB ) and DAB ( 3 , 3′-diaminobenzidine ) staining for detection of superoxide and H2O2 respectively . However , the production of both superoxide and H2O2 of ΔsidN mutant colonies grown on iron-depleted ( DM or DM with BPS ) or iron-replete ( DM supplemented with iron or PD ) media showed no consistent significant difference compared to WT ( results not shown ) . We did on the other hand observe an obvious increase in H2O2 ( by DAB staining ) across all mutant and control colony strains when grown on iron supplemented DM compared to iron-depleted DM alone ( results not shown ) indicating the general influence of free iron supplementation to increase ROS levels .
This study presents the characterization of extracellular siderophore biosynthesis in the foliar grass endophytic fungus E . festucae of the Clavicipitaceae . Through targeted gene disruption of the NRPS gene sidN ( formerly NRPS2 ) and complementation of a mutated ΔsidN strain , we confirmed that sidN encodes a siderophore synthetase required for the production of the novel extracellular siderophore epichloënin A , a variant of ferrirubin of the ferrichrome family . In silico analysis of the SidN domain architecture indicates that the protein consists of three complete A-T-C modules , two terminal T-C repeats and is identical to Type II ferrichrome NRPSs [57] . We can therefore infer that SidN activates three component amino acids required for the biosynthesis of a ferrichrome . This interpretation is consistent with the recent structural characterization of epichloënin A that showed it to be a cyclic octapeptide ferrichrome comprised of three contiguous units of trans-AMHO ( the iron-chelating residues of hydroxamate siderophores [28] , [57] ) , a glutamine and four glycine residues [53] . The third A domain ( A3 ) of SidN has been experimentally confirmed to activate AMHO residues [52] . The other two A domains ( A1 and A2 ) are therefore predicted to activate the two other remaining component amino acids [glutamine ( novel component ) and glycine ( commonly incorporated ) ] of epichloënin A . Investigations into the cellular location of epichloënin A and its ferri-form in the WT fungus grown on iron depleted defined medium ( DM+BPS ) showed that while epichloënin A was predominant in the culture supernatant , ferriepichloënin A was relatively more abundant in the mycelium . These results suggest epichloënin A produced by the fungus is secreted into the extracellular fluids where it binds available iron as ferriepichloënin A which is subsequently taken up by the fungus where the iron is retrieved ( referred to as the “shuttle” mechanism; see Howard [78] for details on siderophore transport mechanisms ) . We are investigating whether ferriepichloënin A could also have an intracellular role , in addition to that of ferricrocin ( putatively encoded by NRPS9 ) , a well-known cellular siderophore which we have found only in mycelial extracts ( L . Johnson et al . , unpublished results ) . Colonization of grass leaves by epichloae endophytes occurs by a unique process described in an intercalary growth model that has been experimentally verified by Christensen et al . [43] . The model illustrates how hyphae in the leaf extension zone , extend by intercalary growth ( non-tip hyphal extension ) at a rate that is synchronized with the expansion and migration of leaf cells . Once the leaf ceases to expand , hyphal growth also ceases , but hyphae remain metabolically active , continuing to be of benefit to their host [79] . This mode of endophytic growth in planta ensures that hyphal growth is synchronized with plant growth . The maintenance and regulation of this lifestyle is pivotal for the mutualistic nature of the associations studied here . We have investigated the effects of loss of production of epichloënin A on the mutualistic association formed between E . festucae with its host grass perennial ryegrass and showed that loss of the extracellular siderophore is deleterious for the maintenance of mutualism . Plants infected with ΔsidN mutants were variably stunted , with altered tiller and root morphology compared to WT . The endophytic hyphae in ΔsidN plants are evidently no longer colonizing only by intercalary hyphal extension but also growing from hyphal tips in an unrestricted and sometimes disorientated manner with respect to the leaf axis , typified by hyper-branching and variable diameter . All of these effects are indicative of a change in the nature of the symbiotic relationship between the fungal endophyte and its host from being mainly mutualistic to detrimental . This study is the first account of the consequences of the loss of an extracellular siderophore in a mutualistic system and indicates that balancing of symbiotic iron homeostasis is an important factor in the maintenance of the mutualistic nature of grass-endophyte associations . Despite the fact that loss of extracellular siderophores sometimes causes microbial pathogens to become less pathogenic , or a mutualistic symbiont to turn pathogenic-like , the commonality in this apparent paradox is that iron uptake appears obligatory for survival of all microbes , whether friend or foe . The symbiotic endophyte studied here is confined to the apoplastic spaces of above ground plant parts; it therefore depends completely on host iron for survival and our results suggest it must scavenge for iron via extracellular siderophore-mediated uptake . The inability of ΔsidN mutants to grow on iron depleted ( DM+BPS ) media indicates epichloënin A biosynthesis contributes significantly to iron assimilation in E . festucae . This is substantiated by the expression analysis of putative components of reductive iron assimilation ( RIA ) that are up-regulated in the ΔsidN infected plants , but appear unable to recompense for the lack of epichloënin A . RIA is therefore a functioning high affinity iron uptake system in E . festucae during in planta vegetative growth , but it does not appear to be the major system required for regulated iron uptake . Accordingly , up-regulation of hapX , the transcription factor that represses iron utilizing proteins in ΔsidN infected plants implies that the ΔsidN mutants require more iron than they are receiving via RIA alone . Another possible explanation is that RIA is bringing in sufficient iron but cellular iron handling processes have become deregulated due to the loss of epichloënin A , which could be acting as a regulatory iron sensor and/or cellular iron store ( since epichloënin A is both secreted outside of the cell ( iron-free ) and also found intracellularly bound to iron ) , and consequently the incoming iron cannot be handled properly and thus utilized effectively . The observed proliferation of ΔsidN mutant hyphae in planta , in which the majority of hyphae ( located in the older parts of the plant ) are abnormal ( frequently devoid of cytoplasm ) and some appear dead ( empty of cytoplasm ) , apart from those extensively colonizing the vasculature which have dense healthy looking cytoplasm suggests that hyphae are growing at a rate greater than can be sustained . The effect of the iron chelator BPS on the distribution of chitin in hyphae of the WT and mutant grown in vitro was also revealing in that it suggests that loss of RIA ( via BPS amendment to the media ) drastically changes chitin distribution and therefore cell polarity . We therefore conclude that changes in iron sensing and regulation are presumably the underlying agents responsible for the abnormal appearance of many mutant fungal hyphae in planta . We observed an interesting increase on the endophyte in planta induced alkaloid ergovaline , in plants infected with ΔsidN mutant strains . This could be as a consequence of iron starvation or may be associated with the extensive hyphal tip growth and branching of fungal hyphae observed in the leaf sheaths of these plants . This pattern of growth is normally restricted to basal meristematic tissues , and a detailed dissection study of N . lolii-infected plants [64] found concentrations of ergovaline to be highly elevated in these tissues . Oide et al . [19] present consequences of extracellular fungal siderophore loss in several fungal phytopathogens through NPS6 gene deletion that resulted in reduced virulence and hypersensitivity to H2O2 [80] , the latter being consistent with our findings of increased H2O2 sensitivity in ΔsidN mutants . The host penetration ability of these Δnps6 strains was not affected , however plant colonization was defective [19] . Application of iron restored this defect indicating that iron deficiency caused the lack of virulence in these fungi . Our work demonstrates that extracellular siderophores play a critical role in fungal-host relationships , but the consequences of loss of siderophore differ depending on the nature of that relationship ( pathogenic vs . mutualistic ) . The endophyte is housed in the apoplastic space where free iron is presumed to be a limiting factor . Our findings indicate that the endophyte requires its extracellular siderophore for mutualistic growth and supports the idea that iron is not readily accessible within the apoplast . The symbioses formed between epichloae endophytes and the Poaceae have co-evolved over evolutionary time , and maintaining symbiotic iron homeostasis is likely to be an intricately balanced process , and a key factor in keeping hyphal growth controlled . Iron homeostasis of the whole endophyte-grass symbiotum must be balanced for two primary reasons . Enough iron must be supplied for metabolism of both plant and fungal partners and excess must be avoided since free iron is toxic due to the formation of reactive oxygen species produced by the Fenton reaction giving rise to oxidative stress [4] . In vitro studies indicate the ΔsidN mutants are significantly more sensitive to H2O2 on iron-depleted medium than on iron-replete medium . This could simply be explained by a reduction in iron-dependent antioxidative enzymes , such as catalases and peroxidases , which require heme . A similar phenomenon was reported for Δnps6 strains where iron application enhanced tolerance to H2O2 and KO2 [19] . Intriguingly , the Fenton reaction does not appear to be the source of oxidative stress sensitivity since iron application would be expected to promote the Fenton reaction . Furthermore , Oide et al . [19] were able to show that both iron-saturated and nonsaturated siderophore ( desferrioxamine ) alleviated reactive oxygen species hypersensitivity indicating a possible protective role for siderophores against reactive oxygen species ( ROS ) in vitro; however their studies to demonstrate this role in planta were inconclusive . Further insights into the connections between ROS and iron regulation have been demonstrated from research into the monothiol glutaredoxins Grx3 and Grx4 of Sacchromyces cerevisiae , which show they function in the defense against oxidative stress through the regulation of iron homeostasis [81] . Research on the NADPH oxidase ( Nox ) complex in E . festucae associations with perennial ryegrass ( through fungal mutants NoxA , NoxB , NoxR , RacA and BemA ) have shown that Nox produced ROS plays a key role in regulating hyphal growth and branching in the host [74]–[77] . Disruption of the spatial and temporal production of fungal ROS leads to the loss of the characteristic features of endophyte-grass mutualistic associations . As the ΔsidN mutant infected plants shared some phenotypic features with the E . festucae Nox mutants , such as hyperbranching of hyphae within host leaves and plant stunting along with increases in tiller number , we explored whether the transcriptional regulation of the E . festucae Nox complex genes was affected in the ΔsidN infected plants . The Nox regulator racA was significantly down-regulated in both ΔsidN mutants studied , and although noxR gene expression was not differentially expressed in the ΔsidN mutants , NoxR requires a functional RacA to spatially regulate ROS production and control hyphal branching [74] , [76] . However , this did not result in changes in ROS production levels in colonies of ΔsidN mutants grown on iron-depleted ( or replete ) medium . Based on these in vitro results , it seems less likely to find a significant difference in ROS production in the ΔsidN mutant infected host plants and was therefore not pursued . Ultimately , to ensure continuance of mutualism , we postulate that maintenance of restricted hyphal growth of E . festucae in planta does not only require a functional Nox complex , but also the maintenance of iron homeostasis which is mediated via epichloënin A . To recapitulate , our results demonstrate that iron acquisition through siderophore-mediated iron uptake is necessary to maintain endophyte mutualism with perennial ryegrass .
Epichloë festucae strain Fl1 ( ex cultivar SR3000 ) and derivatives ( this study ) were grown on 2 . 4% potato dextrose agar ( PDA , Difco Laboratories ) and maintained as previously described [82] , [83] . Defined medium ( DM ) for iron growth studies and chemical analyses were modified from Mantle and Nisbet [84] , with yeast extract replaced with 0 . 6 µM thymine and iron was omitted . DM medium was also supplemented with the following as indicated: 100 µM BPS ( bathophenanthrolinedisulfonic acid; Sigma ) , 20 µM FeSO4 , 20 µM FeCl3 . Inoculation of seedlings of perennial ryegrass ‘Nui’ was performed using the method of Latch and Christensen [85] . Plant growth from seed to 6 weeks-old is as described by Tanaka et al . [86] , except the potting mix was of the following composition: 60% peat , 40% coarse sand with slow release fertiliser ( nutricote with F . T . E . ) and dolomite lime at 2 . 4 kg/m . Determination of endophyte infection was performed by immunoblotting [87]–[89] and light microscopy of epidermal leaf sheaths by aniline blue staining [89] . Plates containing 2% water agar were inoculated with small agar plugs of mycelium and grown for 2 weeks at 22°C , followed by 2 weeks at 4°C . Slides were then mounted with agar blocks using a stereomicroscope to locate appropriate regions for counting of conidia by light microscopy using bright field optics . To determine conidia number from each colony mounted , 5 regions were counted starting from the colony edge and moving towards the colony centre . Conidia numbers were recorded and the data graphed is the mean spore count obtained from three independent colonies and three technical replicates . Sensitivity to oxidative stress was examined using a final concentration of 0 . 7 mM H2O2 in the medium ( DM or PD ) and tests were repeated at least 3 times . The colony diameters obtained from growth on DM or PD with or without H2O2 supplementation were measured and recorded at 7 days ( at 22°C ) . The ratios of DM/DM+H2O2 or PD/PD+H2O2 were recorded for each fungal strain . Data were analysed by an analysis of variance ( ANOVA ) and the least significant difference used to compare stress sensitivity of strains . Plasmid or fosmid DNA was isolated using the QIAprep Spin Miniprep Kit ( Qiagen ) . Fosmid DNA prior to extraction was induced to high copy number using the manufacture's protocol ( Epicentre Biotechnologies ) . Escherichia coli TOPO strain ( Invitrogen ) was used to propagate plasmids using standard techniques [90] . Fungal genomic DNA ( gDNA ) for Southern Blot analysis was isolated using freeze-dried or fresh mycelium by the method of Yoder [91] . For identifying a positive disruption event by PCR , a small scale DNA protocol was used . Transformants were inoculated from one small mycelial plug ( taken from the outer colony margin ) into 100 µl of PD broth in a 1 . 5 ml tube and grown at 26°C for 3 days . Supernatant was removed and freeze-dried mycelia were ground and extracted in 150 µl of lysis buffer ( 100 mM Tris-Cl , pH 8 . 0 , 100 mM EDTA and 1% SDS ) . Following incubation at 70°C for 30 min , the lysates were mixed with 150 µl of 5 M potassium acetate solution and incubated on ice for 10 min . After centrifugation for 10 min , the supernatant containing the fungal DNA was precipitated with 0 . 7 volumes of isopropanol . The DNA pellet was washed once with 70% ethanol and finally dissolved in 20 µl of water . 1 µl of DNA was used for PCR analysis . Isolation of high molecular weight DNA for fosmid library preparation from protoplasts was performed based on the method by Denning et al . [92] and modified as follows: 500 µl of protoplasts in STC buffer were prepared as described by Young et al . [93] , [94] , lysed , and then proteinase K treated and phenol-chloroform purified as stated by Denning et al . [92] . The resulting aqueous phase was precipitated with isopropanol , followed by RNase A treatment ( 0 . 06 mg of RNase A ( Invitrogen ) ; incubation at 37°C for 30 min ) , then precipitated with 100% ethanol , and the resulting pellet washed with 70% ethanol , air dried and finally resuspended in water . For Southern blot analysis , restriction enzyme digested gDNA was transferred to Hybond N+ ( Amersham Pharmacia Biotech ) overnight with 0 . 4 M NaOH . Filters were hybridized at 42°C with digoxigenin ( DIG ) -labelled DNA probes which were labelled by PCR incorporation of DIG-11-dUTP according to the manufacturer's instructions ( Roche ) . Standard PCR conditions for amplification from DNA templates were performed in a 15 µl reaction volume containing 20 mM Tris-HCl ( pH 8 . 0 ) , 50 mM KCl , 1 mM MgCl2 , 200 µM dNTPS , 0 . 3 µM of each primer , 0 . 9 U of Taq DNA polymerase ( Invitrogen ) . Cycling program ( for products less than 2 . 0 kb ) was as follows: one cycle at 95°C for 2 mins; 30 cycles at 95°C for 30 s , 58°C ( dependant on primer TM ) for 30 s and 72°C for 30 s; 72°C for 10 min . An E . festucae fosmid library using gDNA isolated from protoplasts of E . festucae strain FL1 was constructed using the CopyControl Fosmid Production Kit ( Epicentre Biotechnologies ) according to the manufacturer's instructions . Broth cultures from 3840 independent colonies were obtained and a single pooled 384-well plate ( containing 3840 colonies ) was screened by PCR using primers Sid1F & Sid1R . See Table S1 for details of primers used in this study . Standard PCR conditions were used as cited above . We had previously probed a N . lolii Lp19 lambda library with a NRPS2 derived PCR product , and identified one lambda clone of approximately 5 . 8 kb with the following partial domain structure: A ( truncated ) -T-C-A-T-C-T-C-T-C . This information was used for performing a gene disruption by homologous recombination in the sexual relative , E . festucae wild-type ( WT ) strain Fl1 . The gene disruption vector was constructed using Multisite Gateway Three-fragment Vector Construction Kit ( Invitrogen ) so that a portion of the genomic region encoding the third A domain of sidN was replaced with the hygromycin B resistance gene ( see Figure S3A ) . The 5′ and 3′ entry clones were respectively PCR amplified from E . festucae ( primer pairs Sid5F and Sid5R/Sid6F and Sid6R ) , then combined to produce a destination vector containing inserts of approximately 3 . 0 kb of 5′ NRPS coding sequence , followed by a 4 . 0 kb hygromycin B resistant gene ( originally amplified from pAN7-1 ) and approximately 3 . 0 kb of consecutive 3′ NRPS sequence after a deleted region of 35 bp . Protoplasts were prepared as described by Young et al . [93] , [94] . PEG-mediated transformation of protoplasts was carried out with a linear PCR product of ∼10 kb derived from the sidN gene disruption construct and obtained by PCR using the ΔsidN disruption vector as a template with primers sid2F and sid2R . PCR amplification in a 50 µl reaction volume contained 1× Tuning buffer with 5 mM Mg2+ ( Eppendorf ) , 500 µM dNTPs , 400 nM of each primer , 1 ng of plasmid DNA and 2 units of TripleMaster Polymerase Mix ( Eppendorf ) . Cycling program was as follows: one cycle at 93°C for 3 min; 20 cycles at 95°C for 15 s , 56–58°C for 30 s , 68°C for 8 min; 68°C for 10 min . Transformants generated from the gene disruption experiment were initially screened by PCR with primers specific to sidN ( Sid4F and Sid4R ) which flank the hygromycin B resistance gene ( wild-type 0 . 24 kb , disruption 4 . 24 kb; data not shown ) . Southern blot analysis using a probe ( amplified with primers Sid2R and Sid3F ) specific to sidN which also spans the hygromycin B resistance gene in the disruption construct confirmed the disruption event ( mutant strains showed loss of WT 5 . 45 kb band and a new band of 9 . 45 kb ) ( see Figure S3B ) . Putative ΔsidN mutants were subsequently confirmed by southern blotting to have the disruption event ( see Figure S3B ) . Complementation of ΔsidN 85 strain was carried out by co-transforming 5 µg of fosmid DNA [containing the entire promoter and open reading frame of sidN ( as well as 2 additional ORFs located 3′ of sidN ) ] along with 1 µg of circular PII99 [95] carrying the selectable antibiotic resistant marker geneticin . The transformation methodology of Vollmer and Yanofsky [96] with modifications by Itoh et al . [97] was used for gene disruption and complementation experiments . Disruption transformants were selected on regeneration ( RG ) medium containing hygromycin ( 150 µg ml−1 ) , whereas geneticin resistant transformants derived from the complementation experiment were selected on RG medium containing 200 µg ml−1 of geneticin . Mycelium was subcultured three times for nuclear purification of transformants . Complementation transformants were identified by their ability to grow on iron-depleted medium containing the iron chelator BPS . Confirmation was obtained by assaying culture filtrates from a subset of positive transformants for siderophore production by LCMSMS . qPCR on gDNA isolated from endophyte-infected plants ( pseudostem ) was performed on a MyiQ cycler ( Bio-Rad ) using primers designed to a nonribosomal peptide synthetase ( NRPS-1 ) gene as described in Rasmussen et al . [60] . Concentration of endophyte in infected tissues is expressed as the number of copies of the single copy NRPS1 gene per total gDNA ( from plant and fungus ) . Total RNA was extracted from frozen fungal mycelium using TRIzol reagent ( Invitrogen ) . For extraction from perennial ryegrass tissues , RNA was at first isolated using TRIzol and further purified through the Plant RNeasy Kit ( Qiagen ) as follows: ∼30 mg of frozen plant tissue was ground to a fine powder , mixed well with 750 µl of TRIzol and incubated at room temperature for 5 min . Following centrifugation at 12K rpm for 10 min , the supernatant was transferred to a new tube and 150 µl of chloroform added . After mixing well for 15 s and incubating at room temperature for 3 min , samples were centrifuged at 13 . 2 K rpm for 5 min . The aqueous phase was transferred to a fresh tube and an equal volume of 70% ethanol added before loading onto a Qiagen RNeasy mini column ( pink ) . From this point onwards , total RNA was column purified using the manufacturer's instructions ( Qiagen ) . The optional DNase digestion on the column step was also carried out for plant samples . Any DNA still remaining was removed from total RNA by treating 10 µg of RNA with 20 U of DNase I , RNase-free ( Roche ) and 5 mM MgSO4 at 37°C for 30 min , followed by 5 min at 75°C . First strand cDNA primed with oligo ( dT ) was synthesized using the ThermoScript RT-PCR system ( Invitrogen ) from 1 or 2 µg of denatured total RNA by incubation at 50°C for 60 min , followed by 85°C for 5 min . For RT-PCR expression analysis of sidN from axenic liquid DM cultures , the primer pair Sid4F and Sid4R was used to amplify cDNA using standard PCR conditions . Primers to the Neoptyphodium lolii actin gene ( Acting and ActinR ) were used to check for cDNA quality . Two-step RT-qPCR analysis of perennial ryegrass tissue was performed on an iCycler ( MyiQ Single Color Real-Time PCR Detection System , Bio-Rad ) using Power SYBR Green PCR Master Mix ( AB applied biosystems ) . The following cycle parameters were applied: 95°C for 5 min and then 40 cycles of 95°C for 20 s , 54–56°C for 20 s , and 72°C for 30 s followed by a melt curve analysis . Primer efficiency ranged between 94% and 110% . Primers for RT-qPCR ( see Table S1 ) of putative iron-regulated genes , Ftr1 , Fet3 and HapX were designed to sequences from the E . festucae genome strain E2368 . These genes were identified by performing tblastx queries of the E . festucae ( E2368 ) genome with characterized iron-regulated genes from other fungal species ( http://csbio-l . csr . uky . edu/ef2011/blast/blast . html ) . The deduced amino acid sequences of the identified genes from E . festucae ( strain E2368 ) were then compared to the NCBI protein database by blastx ( see Table S2 for descriptions of identified genes used for RT-qPCR ) . Primers for the RT-qPCR analysis of NADPH Oxidase ( Nox ) genes ( see Table S1 ) were designed to E . festucae sequences sourced from accession numbers: AB236860 ( noxA ) , AB236861 ( noxB ) , AB260938 ( noxR ) and AB260937 ( racA ) . Purified PCR products were used to create a calibration curve for each gene-specific primer pair . Two endophyte housekeeping genes ( a 60S ribosomal protein L35 and gamma actin , see Table S1 for details ) were used to normalize the expression levels of the different genes . An analysis of variance ( ANOVA ) was performed from the geometric mean obtained from three biological replicates and two or three technical replicates for each primer pair . The least significant difference was used to compare the samples . P-values obtained for each transcript were: ftrA P<0 . 001; fetC , P<0 . 001; hapX P<0 . 002 , noxA P<0 . 001 , noxB P = 0 . 005 , noxR P = 0 . 017; racA P<0 . 001 . Sequencing was performed with the Big-Dye ( Version 3 ) chemistry ( PE biosystems ) and the products separated on an ABI Prism 3100 automated sequencer ( Applied Biosystems ) . Sequence comparisons were performed against local mirrors of a number of public databases , including GenBank , RefSeq , Cogeme and a selection of fungal genomes from the Broad Institute . Algorithms tblastx , blastx and blastn were employed to generate alignments [98] . Contigs were assembled using Vector NTI Advance 9 . 1 . 0 , ContigExpress ( Invitrogen ) and SEQUENCER 4 . 6 ( Gene Codes Corporation ) . NRPS domain structure was determined using a local mirror of InterProScan combined with a manual annotation based on the identification of motifs within domains as described by Schwarzer et al . [99] . Samples of supernatant from liquid cultures grown for 2 weeks were separated for analysis by centrifugation . The residual mycelium was freeze-dried and finely ground , and extracted with water , and siderophores separated by solid phase extraction [100] . All samples were stored at −20°C prior to analysis . Milli-Q water and HPLC grade solvents were used for LCMS . The samples were thawed prior to analysis and transferred to a HPLC vial with 200 µl insert . Samples were kept at 5°C in the autosampler , and 10 µl subsamples were injected . Analytes were eluted through a C18 Luna column ( Phenomenex Torrence , CA , USA ) ( 150×2 mm , 5 µm ) at a flow rate of 200 µl min−1 using a Thermo Finnigan Surveyor HPLC system with a solvent gradient ( solvent A: H2O 0 . 1% formic acid; B: MeCN 0 . 1% formic acid ) , starting with 5% B , 95% A for 5 min and then increasing to 33% B after 15 min , then to 95% B by 20 min where the composition was held for 5 min to wash the column before being returned to 5% B to re-equilibrate the column . Mass spectra were detected with a linear ion trap mass spectrometer ( Thermo LTQ ) using ESI in positive ion mode . The spray voltage was 4 . 5 kV and the capillary temperature 275°C . The flow rates of nitrogen sheath gas , auxiliary gas , and sweep gas were set to 20 , 5 , and 10 ( arbitrary units ) , respectively . Ferriepichloënin A was detected as an MS1 ion of m/z 569 [MH2]2+; epichloënin A was detected as an MS1 ion of m/z 542 [MH2]2+ . Cis-AMHO in a standard solution ( kindly provided by T . Verne Lee , University of Auckland , New Zealand ) and trans-AMHO in culture supernatants were detected by monitoring the MS1 ion of m/z 261 , and MS2 spectra were recorded by selecting and fragmenting this ion . Guttation fluid was collected from the association of L . perenne G1057 with E . festucae Fl1 , as previously described by Koulman et al . [61] . In brief , plants were placed overnight in a closed container and in the early morning the fluid accumulated at the leaf ends of a plant was collected with a glass pipette , transferred to a plastic container , and stored at −20°C until analysis . Samples were analysed by direct injection of a 10 µl subsample of guttation fluid into a Thermo Finnigan Surveyor HPLC system attached to a linear ion trap mass spectrometer ( Thermo LTQ ) operated as described above . Ferriepichloënin A was detected by selecting and fragmenting the parent [MH2]2+ ion m/z 569±2 ( 35% relative collision energy ) , and selecting and fragmenting the product ion m/z 1024±2 ( 35% relative collision energy ) and monitoring the total ion current . HPLC was used to measure in planta levels of lolitrem B , ergovaline and peramine alkaloids as previously described [101] , [102] . Hyphae within leaf sheaths were examined by light microscopy of epidermal leaf sheaths by aniline blue staining [89] . For calcofluor ( 1 µg/ml ) staining , mycelia were grown on microscopy slides covered with a thin layer of DM medium . Samples were directly incubated with the dye for up to 20 minutes at room temperature and washed with water to remove background fluorescence . Superoxide production was detected by NTB staining using the method described by Takemoto et al . [74] except slides were overlaid with DM or PD media . H2O2 production was examined by DAB staining ( DAB forms a brick-red precipitate upon reaction with H2O2 [103] of cultures grown on DM and PD media as described by [104] . Examination of sections from pseudostem samples ( approximately 1 mm long ) required fixation in 3% glutaraldehyde and 2% formaldehyde in 0 . 1 M phosphate buffer pH 7 . 2 for 2 h at room temperature , followed by treatment in 1% osmium tetroxide in 0 . 1 M phosphate buffer pH 7 . 2 for 0 . 5 h at room temperature . The tissues were then washed 3 times in 0 . 1 M phosphate buffer pH 7 . 2 , dehydrated in an acetone/water series and two times in 100% acetone . Samples were infiltrated with an acetone/polarbed 812 resin mixture ( 50/50 , v/v ) , and then embedded in fresh resin mixture in silicone rubber moulds and cured for 48 h at 60°C . To examine the distribution of hyphae within different ages of leaf sheaths , cross sections ( approximately 1 µm thick ) were prepared from the treated pseudostem samples , and stained with toluidine blue for examination by bright field light microscopy . For examination with a Philips 201C transmission electron microscope , ultra-thin sections were cut and stained with saturated uranyl acetate in 50% ethanol for 4 min followed by lead citrate for 4 min . Sequence data from this article can be found in the GenBank/EMBL databases under the following accession numbers: N . lolii NRPS2 ( EF19536 ) , N . lolii ( strain Lp19 ) sidN ( JN132404 ) , E . festucae ( strain Fl1 ) sidN ( JN132407 ) , E . festucae ( strain E2368 ) sidN ( JN132403 ) , E . festucae ( strain E2368 ) ftrA ( JN132405 ) , E . festucae ( strain E2368 ) fetC ( JN132406 ) , E . festucae ( strain E2368 ) hapX ( JN132401 ) , E . festucae ( strain E2368 ) actG ( FJ826616 . 1 ) , E . festucae ( strain E2368 ) actA ( FJ379533 . 1 ) , N . lolii RP ( L35 ) ( JN132402 ) . | Perennial ryegrass is a cool-season grass that is agriculturally important as forage , especially in Australasia . Essential to the longevity and maintenance of healthy pastures are epichloae fungi , such as Epichloë/Neotyphodium spp . that live inside these grasses in a symbiotic manner . These fungi produce bioactive compounds that can protect its host grass from various abiotic and biotic stresses . They depend on resource supply by their host , including nutrients that are taken up by the roots of the grass host . We have found that iron , an indispensable element for growth , is acquired by Epichloë festucae from its grass host via the secretion of the extracellular siderophore epichloënin A . Without iron-mediated siderophore uptake to capture host iron , the symbiotic relationship becomes disturbed and the growth habit of the fungus is no longer restricted . Our results indicate that secretion of epichloënin A enables the symbiotic fungus to compete for plant iron while avoiding fungal over-growth inside the host plant . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genetically",
"modified",
"organisms",
"gene",
"identification",
"and",
"analysis",
"genetics",
"molecular",
"genetics",
"biology",
"agricultural",
"biotechnology",
"agriculture",
"gene",
"function"
] | 2013 | An Extracellular Siderophore Is Required to Maintain the Mutualistic Interaction of Epichloë festucae with Lolium perenne |
Variation in the TGF-β signaling pathway is emerging as an important mechanism by which gonadal sex determination is controlled in teleosts . Here we show that amhy , a Y-specific duplicate of the anti-Müllerian hormone ( amh ) gene , induces male sex determination in Nile tilapia . amhy is a tandem duplicate located immediately downstream of amhΔ-y on the Y chromosome . The coding sequence of amhy was identical to the X-linked amh ( amh ) except a missense SNP ( C/T ) which changes an amino acid ( Ser/Leu92 ) in the N-terminal region . amhy lacks 5608 bp of promoter sequence that is found in the X-linked amh homolog . The amhΔ-y contains several insertions and deletions in the promoter region , and even a 5 bp insertion in exonVI that results in a premature stop codon and thus a truncated protein product lacking the TGF-β binding domain . Both amhy and amhΔ-y expression is restricted to XY gonads from 5 days after hatching ( dah ) onwards . CRISPR/Cas9 knockout of amhy in XY fish resulted in male to female sex reversal , while mutation of amhΔ-y alone could not . In contrast , overexpression of Amhy in XX fish , using a fosmid transgene that carries the amhy/amhΔ-y haplotype or a vector containing amhy ORF under the control of CMV promoter , resulted in female to male sex reversal , while overexpression of AmhΔ-y alone in XX fish could not . Knockout of the anti-Müllerian hormone receptor type II ( amhrII ) in XY fish also resulted in 100% complete male to female sex reversal . Taken together , these results strongly suggest that the duplicated amhy with a missense SNP is the candidate sex determining gene and amhy/amhrII signal is essential for male sex determination in Nile tilapia . These findings highlight the conserved roles of TGF-β signaling pathway in fish sex determination .
Master sex-determining ( SD ) genes are the key genetic switches controlling the gonadal sex differentiation cascade leading to the development of either ovaries or testes . To date , master SD genes have been identified in only a few vertebrate species . SRY/Sry was the first sex determiner identified in mammals [1 , 2] . With the recent discovery that sox3Y is the sex determiner in Oryzias dancena [3] , Sox genes continue to figure prominently in discussions of vertebrate sex determination . Doublesex/mab ( DM ) related genes have been associated with sex determination in a wide range of species , including Dmrt1 in chicken and half-smooth tongue sole [4 , 5] , DM-W in African clawed frog [6] , and dmy/dmrt1bY in Oryzias latipes [7 , 8] . Other genes have been implicated as master sex determiners in particular lineages , including FOXL2 in goat [9] , and sdY ( irf9y ) in rainbow trout [10] . Several recent studies have suggested that components of the transforming growth factor beta ( TGF-β ) signaling pathway are involved in sex determination in fishes . These include a Y-linked duplicate of the anti-Müllerian hormone ( amhy ) in the Patagonian pejerrey [11] , a mutation in the amh receptor ( amhrII ) in Takifugu rubripes [12] , and a Y-linked duplicate of a related ligand , gonadal soma derived growth factor ( gsdfY ) in Oryzias luzonensis [13] . These findings suggest a critical role for TGF-β signaling in gonadal sex determination in teleosts . Studies of mammalian sex chromosomes have provided significant insights into the evolution of sex determination , but SD genes have not yet been identified in the vast majority of vertebrates . For example , teleost fishes make up nearly half of all living vertebrate species and show a wide variety of sex determination mechanisms [14] , but only a handful of these sex determiners have been identified . Closely related species of fish frequently segregate different master sex determiners , suggesting that a delicately balanced network of gene interactions controls sex determination . For example , three different genes ( dmy , gsdfY and sox3Y ) have been identified as master sex determiners among closely related species of ricefish [3 , 7 , 8 , 13] . Master sex determiners map to three different chromosomes among closely related species of stickleback [15] . Recent work has identified at least three sex determiners among strains of zebrafish [16 , 17] . Numerous studies have investigated the mechanisms of sex determination in tilapia ( Oreochromis niloticus ) , motivated in part by commercial interest in the higher growth rates of all-male progenies . Tilapia are gonochoristic teleosts in which sex is largely genetically determined [18] , although environmental factors also play a role [19] . XX/XY sex determination systems have been described on both LG1 and LG23 in this species [20 , 21] . All-XX and all-XY progenies can be obtained by crossing normal XX females to either experimentally sex-reversed XX pseudomales or YY supermales respectively [22] . A previous report identified several sex-linked markers near the amh gene on LG23 [21] . More recent studies identified a Y-linked duplication of amh on LG23 , termed a male-specific amhy , differing from the sequence of amh by a 233 bp deletion in exonVII [23] . Our own analyses have identified five additional sex-linked markers on LG23 that map very close to amh [22] . amh is located in the center of this sex-linked region and shows sexually dimorphic expression in the gonads at 3 days post fertilization [24] , making it an interesting gene for sex determination in this species . Amh is responsible for the regression of Müllerian ducts in tetrapods [25] . It is also found in teleost fish despite the fact that they do not have Müllerian ducts [23 , 26–28] . In mammals , Amh functions primarily through the type II receptor AmhrII [25] . Mutations of amhrII in medaka and Takifugu rubripes result in male to female sex reversal [12 , 29] . These studies suggested that amh/amhrII signaling might play a role in fish sex determination . Recent efforts have generated a number of important resources for tilapia research , including a genome sequence , a microarrayed fosmid library , and several gonadal transcriptomes [30–32] . TALEN and CRISPR/Cas9 gene knockout technologies have also been established in tilapia [33 , 34] . The availability of these tools prompted us to try to isolate the SD gene in the Nile tilapia . In the present study , we isolated a Y-specific duplicate of the amh gene , designated as amhy , and confirmed its male-specific ( XY and YY ) expression by transcriptome analysis and Western blot . We used transgenic techniques to overexpress amhy in XX fish , and we used CRISPR/Cas9 mutagenesis to knockout amhy , and its putative receptor amhrII in XY fish . Our results suggest a conserved role for the TGF-βsignaling pathway in sex determination of vertebrates .
We used PCR to screen a microarrayed tilapia XY genomic library for a sex-specific marker . We identified X-specific and Y-specific fosmid clones ( designated X278 and Y156 ) producing amplimers of 1422 and 982 bp , respectively ( S1 Fig ) . The two fosmid clones were then sequenced using Illumina HiSeq2000 technology and carefully assembled using local Basic Local Alignment Search Tool ( local BLAST ) by hand based on the sequence differences of several genomic PCR products amplified from XX and YY individuals and the sequence differences between Y156 and X278 fosmid . Sequence analysis revealed that an amh gene , termed as amh , was present in the X278 fosmid clone . By gene prediction with GENSCAN and BLAST search in the tilapia genome using the assembled sequence of the Y 156 clone , only two genes , designated as amhy and amhΔ-y , were found in the Y156 fosmid clone . amhy is a tandem duplicate located immediately downstream of amhΔ-y ( Fig 1a ) . The insertion of Y156 was about 40 kb , which was further confirmed by sequencing 25 fragments , each about 3 kb with partial overlapping ends . A comparison of the sequences of the two clones revealed numerous differences . The coding sequence of amhy was identical to the X-linked amh except a single nucleotide polymorphisms ( SNP ) ( C/T ) in exonII , which changes an amino acid ( Ser/Leu92 ) in the N-terminal region . amhy has lost 5608 bp of promoter sequence that is found in the X-linked amh homolog ( Fig 1a and 1b ) . amhΔ-y contains a 5 bp ( ATGTC ) insertion in exonVI , resulting in a frameshift mutation and a premature stop codon . The 5 bp insertion could be easily confirmed by Taqα I restriction enzyme digestion of the genomic PCR fragments spanning it , which resulted in two digested bands in the XY and YY fish , whereas no band was produced in XX fish ( S2 Fig ) . amhΔ-y also has a deletion of 233 bp in exonVII that further precludes translation of the protein motif that binds to the TGF-βreceptor . There was one additional insertion ( 36 bp ) and three other deletions ( 4 , 472 and 161 bp , respectively ) at -1972 , -1756 , -1664 and -625 , respectively , of amhΔ-y start codon ( ATG , A as 0 point ) , compared with amh on the X chromosome ( Fig 1a ) . These differences between amhy , amh and amhΔ-y were further demonstrated by PCR amplification in XX , XY and YY genomic DNA pools ( S3 Fig ) . The gene sequences of amhy , amhΔ-y and amh are shown in the S1 Sequence . Primers flanking differentiated sequences in the promoter and exon region were used for PCR-based sex genotyping . Y-specific primers amplified fragments in the XY and YY pooled genomic DNA respectively , but not in the pooled XX genomic DNA ( Fig 1c ) . Genotypic sex based on these primers showed 100% correspondence with phenotypic sex based on gonadal histology in 300 individuals derived from four crosses ( S1 Table ) . Finally , there are five SNPs in the coding sequences of amhΔ-y and amhy . Three of them are non-synonymous and the amino acid changed ( S4 Fig ) . amhΔ-y is close to neutral evolution in comparison to amhy ( Ka/Ks<1 ) ( S2 Table ) . The tilapia amhΔ-y cDNA isolated by rapid amplification of cDNA ends ( RACE ) is 2 , 065 bp long with a 46 bp 5' untranslated region ( UTR ) , a 1 , 242 bp 3' UTR and an ORF of 777 bp encoding a putative protein of 258 aa ( amino acid ) without the TGF-β domain . The amh/amhy isolated by RACE has a 346 bp 5' UTR , a 782 bp 3' UTR and an ORF of 1 , 545 bp , encoding a putative protein of 514 aa with the TGF-β domain . RT-PCR showed that among the twelve tissues examined , amh/amhy and amhrII were expressed exclusively in gonads , with greater expression in testis . amhΔ-y was expressed only in the XY testis , not in the XX ovary ( S5 Fig ) . Transcriptome analysis revealed that amhy and amhΔ-y transcripts were only detected in XY gonads , with expression at 5 days after hatching ( dah ) , the critical period for molecular sex determination in Nile tilapia , peaked at approximately 30 dah , and decreased at 90 and to very low level at 180 dah ( Fig 2a and 2b ) . Similar expression profiles of amh and amhrII were observed in both XX and XY gonads , with significantly higher expression in XX than in XY gonads at 5 dah , while they showed higher expression in XY than in XX gonads at 30 dah onwards ( Fig 2c and S6 Fig ) . Specificity of the Amh ployclonal antibody was characterized by Western blot analysis using recombinant protein ( both purified and unpurified ) , and the proteins extracted from the XX ovary , XY testis , and YY testis . The Amh antibody can recognize the Amh , AmhΔ-y and Amhy protein . The blots revealed specific bands of ~54 kDa , corresponding to the calculated molecular weight of Amh/Amhy , in the protein samples extracted from the XX ovary , XY and YY testis . Another band of ~27 kDa , corresponding to AmhΔ-y , was detected exclusively in the protein samples extracted from the XY and YY testis . Amh/Amhy was detected in XY gonads of 5 , 10 , 30 , 90 , 180 and 300 dah tilapia , and also with expression of Amh in XX tilapia at these stages , while AmhΔ-y was only detected in 5 , 10 , 30 and 90 dah XY gonads . In addition , Amhy and AmhΔ-y were also detected in the proteins from YY gonads of 30 and 90 dah tilapia ( S7 Fig ) . By immunohistochemistry , the Amh/AmhΔ-y/Amhy and AmhrII proteins were located in somatic cells surrounding germ cells in the XY gonads at 5 dah . At later stages , Amh/AmhΔ-y/Amhy is observed in myoid cells and Sertoli cells , while AmhrII was observed in spermatogonia and Sertoli cells of the testis at 30 , 90 and 180 dah ( S8 Fig ) . Two guide RNAs ( gRNA ) were designed , in exons II and III of amhy/amh/amhΔ-y , to increase the likelihood of successful targeting and also exclude the effects of off-target events on the phenotypes . The gRNAs contained BstN I or BsrB I sites adjacent to the protospacer adjacent motif ( PAM ) sequence for mutation analysis . Restriction enzyme digestion and Sanger sequencing were performed to confirm the insertions or deletions ( indels ) , including both in-frame and frame-shift mutations , in pools of randomly selected embryos ( Fig 3a ) . The screening results showed that 18% ( 8/45 ) of target 1 and 62% ( 28/45 ) of target 2 fish were mutated ( S3 Table ) . Due to the high homology of amhy , amhΔ-y and amh , CRISPR/Cas9 disrupted three genes at both target sites . Restriction enzyme digestion showed that the CRISPR system disrupt amhy and amhΔ-y genes equally in F0 knockout XY fish ( S9 Fig ) . RT-PCR with forward primers located in the targets using cDNA from 90 dah mutated gonads as templates demonstrated that the expression of amh/amhΔ-y/amhy mRNA in the F0 knockout group was much lower than that of control group ( S10A Fig ) . Consistent with this , dramatically decreased Amh/Amhy protein levels and no AmhΔ-y were detected in the gonads of F0 knockout fish by Western blot ( S10B Fig ) . The gonads of the mutagenized fish were subjected to both histological and immunohistochemical ( IHC ) analyses at 10 , 30 and 90 dah . Macroscopic observation of the gonads of 3-month-old gRNA/Cas9 microinjected XY fish revealed that some of the mutated F0 fish showed male to female sex reversal . Gonadal sex differentiation in the sex-reversed XY fish was characterized by the formation of the ovarian cavity ( OC ) and the appearance of phase II oocytes . These gonads were topologically indistinguishable from the control XX ovary , but obviously different from the control XY testis ( Fig 3b ) . IHC revealed that like the control ovary , male Sertoli cell marker Dmrt1 was not expressed in the sex-reversed XY gonads ( Fig 3bA–3bD ) . Additionally , female specific marker Cyp19a1a was expressed in these sex-reversed XY gonads , like the control ovary ( Fig 3bE–3bH ) . Consistent with the IHC results , higher serum estrogen ( E2 ) was observed in the sex-reversed XY fish compared with the XY control ( Fig 3c ) . The effects of amh/amhΔ-y/amhy deficiency on sex determination could be detected from a very early stage by IHC analysis of sexual dimorphic proteins such as Dmrt1 and Cyp19a1a . Tracing back , Cyp19a1a , which was not expressed in XY gonads at 10 and 30 dah , was detected in the sex-reversed XY gonads as early as 10 and 30 dah ( Fig 3bJ , 3bN , 3bL and 3bP ) . In contrast , Dmrt1 , which was detected in the control XY but not in XX gonad ( Fig 3bK and 3bO ) , was not detected in sex-reversed XY gonads ( Fig 3bI and 3bM ) . Both target sites resulted in sex reversal . Of the 8 F0 fish with mutations in target site 1 , 5 ( 62% ) displayed complete male to female sex reversal . Of the 28 F0 fish with mutations in target site 2 , 20 ( 71% ) showed complete sex reversal ( Table 1 ) . The genetic sex of these sex-reversed XY fish with amh/amhΔ-y/amhy mutation was confirmed by amplification of a sex-linked marker ( SPF/SPR ) and all of these sex-reversed males were XY fish ( S11 Fig ) . In addition , a number of oogonia , as demonstrated by Gsdf staining , and only a few primary and secondary oocytes , were observed in Amh deficient XX ovary at 90 dah ( S12A Fig ) , compared with the control ( S12B Fig ) . Sperm from the XY founders carrying different types of amh/amhΔ-y/amhy mutations ( target 1 and 2 ) was mated with wild-type XX fish to produce F1 animals and independent mutant alleles having indels at the target of F1 XY were successfully obtained . These mutants caused frame shifts resulting in premature termination , predicted to yield truncated proteins ( S13 Fig ) . The F1 XY fish with amhy mutant allele on the Y chromosome alone ( n = 9+5 ) or double mutation of amhy and amhΔ-y ( n = 14+10 ) displayed sex reversal with clear ovarian structure at 90 dah . In contrast , the F1 XY fish with mutation of amhΔ-y alone displayed no sex reversal with normal testis at 90 dah ( n = 11+7 ) ( Fig 4 , Table 2 ) . Gonad histological analysis revealed that 8 ( 32% ) of 25 Y156 fosmid transgenic F0 XX fish displayed complete female to male sex reversal at 90 dah ( Fig 5a , Table 1 ) . The sex-reversed gonads exhibited a clear testicular structure . IHC showed that they expressed the testicular specific gene Dmrt1 , and did not express the ovarian specific Cyp19a1a ( Fig 5aA–5aF ) . The integration and mRNA expression of the transgene in the XX fry were examined and confirmed by genomic PCR and RT-PCR , using amhΔ-y specific primers , at 3 and 5 dah ( Fig 5b ) . To investigate the ability of the amhy ORF to induce sex reversal , we constructed overexpression vector pIRES-hrGFP-1a in which CMV controlled the amhy cDNA . Morphologically , the amhy-transgene XX F0 fish also displayed sex reversal with a clear testicular structure ( n = 9 ) , while no sex reversal was found in the control XX group ( n = 21 ) at 90 dah ( Fig 5c ) . In contrast , the amh or amhΔ-y-transgene XX F0 fish displayed no sex reversal with a clear ovarian structure at 90 dah ( amh , n = 11; amhΔ-y , n = 10 ) ( S14 Fig , Table 1 ) . Two gRNAs were designed in the exon II and III of amhrII ( Fig 6a ) , an autosomal gene located on LG20 . In total , 73% ( 22/30 ) of target 1 and 50% ( 15/30 ) of target 2 were mutated by CRISPR/Cas9 . The mutation frequencies within individuals ranged from 22% to 68% ( S3 Table ) . All fish with amhrII mutations showed complete male to female sex reversal , as visualized by the formation of the ovarian cavity ( OC ) and the appearance of pre-vitellogenic oocytes when checked by gonad histology at 90 dah ( Fig 6b ) . IHC analysis showed that Cyp19a1a was expressed like the control ovary , while Dmrt1 was not expressed in these sex-reversed XY gonads ( Fig 6bA–6bD ) . The effect of AmhrII deficiency on gonadal differentiation could be detected from a very early stage . Cyp19a1a was expressed , while Dmrt1 was not expressed in AmhrII deficient XY gonads even at 10 and 30 dah ( Fig 6bE–6bL ) . Consistently , knockout of amhrII in the XY fish resulted in increased serum E2 level compared with the control fish ( Fig 6c ) . The genetic sex ( XY ) of these sex-reversed fish was further confirmed by PCR of a sex-linked marker ( SPF/SPR ) ( Fig 6d ) .
In recent years , several different master SD genes have been identified in various fish species , giving the impression that the molecular mechanism underlying sex determination is different in each group . However , the limited data available for teleost SD genes suggest that members of the TGF-β superfamily ( gsdfY , amhy and amhrII ) could be part of a common pathway for sex determination in fish [11–13] . Tilapia sex determination has been widely studied with the goal of producing all-male progenies that have an enhanced growth rate in aquaculture . In Nile tilapia , some studies using fish from Egypt and Ghana have indicated that the main sex determining locus is on LG1 ( with unresolved evidence regarding LG3 ) [35–37] . Recently , a sex-determining locus was mapped adjacent to the amh on LG23 in the Swansea strain of this species , using Simple Sequence Repeats ( SSR ) and sex specific markers [21 , 24] . Thereafter , amhy , a Y-specific duplicate of the amh gene , was identified and suggested to be the candidate sex determiner [23] . However , this gene , named as amhΔ-y by us , is not responsible for sex determination in our strain . Another duplicated copy of amh with a missense SNP and a large fragment of promoter loss , which is located immediately downstream of amhΔ-y and designated as amhy , was found and demonstrated to be critical for male sex determination in our strain . Therefore , it is possible that different strain possesses different sex chromosome and different SD genes . Judging from the assembled sequence , it is reasonable to conclude that amhy arose from duplication of amh gene followed by 5608 bp promoter loss . amhΔ-y is an allele of the X-linked amh as most of the sequences in the upstream promoter of amhΔ-y were identical to the upstream of amh on the X chromosome . amhy could be considered a candidate SD gene of the Nile tilapia because of its Y specific expression profile . Due to the high homology of three amh , four pairs of gonadal transcriptomes from different development stages were used to analyze their expression . It was found that amhy expression was restricted to XY testis , with expression at the beginning of critical sex determination period 5 dah , clearly preceding the first signs of morphological differentiation of ovaries and testes ( ~25 dah ) [38] . According to the transcriptome data , both amhy and amh mRNA were detected in the XY gonads at 5 dah . Therefore , Western blot results indirectly suggested the expression of Amhy protein in XY gonads at 5 dah . In contrast , amh expression in XY gonads was much lower than that of XX gonads at 5 dah . Although the amhΔ-y was also found to be restricted to XY testis , we found there is a 5 bp insertion in exonVI of the amhΔ-y in addition to the previously reported TGF-β domain 233 bp deletion in exonVII [23] . It is this frameshift mutation that generates a truncated Amh lacking the TGF-β domain , which is important for binding to AmhrII . The truncated Amh could not directly bind to AmhrII even if the 233 bp sequence in exonVII was not deleted . Therefore , amhΔ-y might be a degenerated gene in tilapia . The best way to understand the function of a gene in sex determination is to perform gain/loss of function studies and to characterize the resulting biological effects . For instance , overexpression of a 117 kb genomic DNA fragment that carries dmy in XX medaka , or the presence of a genomic fragment that included gsdfY , converts XX individuals into sex-reversed XX males [13 , 39] . Knockdown of the Patagonian pejerrey amhy in XY fish led to an up-regulation of female factors and the development of ovaries [11] . In the present study , knockout of amhy or both amhy and amhΔ-y using CRISPR/Cas9 resulted in ovarian development in XY fish , while mutation of amhΔ-y alone could not . In contrast , overexpression from the amhy genomic region or its ORF under the control of CMV promoter induced testicular differentiation in XX fish , while overexpression of amhΔ-y alone in XX fish resulted in no sex reversal . Therefore , we demonstrated that amhy is necessary and sufficient to induce testicular differentiation in tilapia . A detailed analysis showed that the missense SNP and the large fragment promoter loss in amhy might contribute to sex determination in tilapia . Recent studies reported that a few regulatory or coding sequence mutations in other pre-existing and duplicated genes can generate new SD genes in fishes . For example , a few differences in the cis-regulatory region of gsdfY and sox3Y contribute to male sex determination in two medaka species . A missense SNP in AmhrII is the only difference associated with phenotypic sex in Takifugu rubripes . In this study , overexpression of amhy ORF in XX fish led to sex reversal , while overexpression of amh in XX fish resulted in no sex reversal , implying that the missense SNP ( C/T ) in the coding sequence might contribute to male sex determination in tilapia . In another Nile tilapia strain , a missense SNP in exonVI of amh was also associated with autosomal and temperature dependent sex reversal [40] . These studies indicated that small variations in the coding sequence of amh might have taken over a critical role in tilapia sex determination . The important characteristic of a master SD gene is its tight linkage with the non-recombinant part of the heterochromosome . Right now , we do not know the size of the male-specific non-recombining region and how many genes are located in the region in tilapia . Therefore , amhy is considered as the candidate sex determining gene in this strain of the Nile tilapia . In this report , an XY specific up-regulation was detected in the expression of amhrII in the gonads from 5 dah onwards , coincident with sex determination in tilapia . Higher levels of amhrII expression in testis have been consistently observed in four tilapia species and several other fishes [12 , 28 , 41] . Importantly , mutations in amhrII in medaka and Takifugu rubripes lead to male to female sex reversal [12 , 29] . In the present study , knockout of amhrII in XY fish resulted in 100% male to female sex reversal in tilapia . However , as in medaka , but unlike in Takifugu rubripes , amhrII is an autosomal gene located on LG20 in Nile tilapia . It is well documented that AmhrII is the receptor for Amh in mammals [25] . However , the receptors for amhy and amh in fish have not been identified . Notably , the expression profiles of amh and amhy was similar to that of amhrII during tilapia gonadal development and these factors were found to be co-localized in cells surrounding the germ cells at 5 dah . Both Amhy and Amh have the identical TGF-β domain , which is responsible for binding to its receptor . According to these results , Amhy might be the ligand of AmhrII . Therefore , Amhy signal functions through AmhrII to determine sex determination in tilapia . Further investigations may reveal the mechanism by which Amhy/AmhrII signal pathway determines tilapia male sex . In addition , knockout of amhrII resulted in 100% sex reversal , while knockout of amhy in F0 fish XY only resulted in about 60% sex reversal . As the F0 fish were mosaic , the mutation rate varies individually . Therefore , only some of the F0 knockout fish displayed sex reversal . A thorough analysis of gene mutation of F1 knockout fish indicated that the ratio for fish bearing amhy frameshift mutation versus fish bearing frameshift mutation is approximately 60% ( 23/34 ) , which is exactly the ratio of sex reversed fish to F1 positive fish . The probability for amhy and amhΔ-y mutation in different cell types of F0 fish should be the same as that of the germ cells . This explains why only 60% of the F0 positive fish displayed male to female sex reversal . In contrast , knockout of amhrII is equal to disrupt whole signal pathway , and therefore , resulted in 100% sex reversal . amhy was first reported as the sex determiner in Patagonian pejerrey [11] . Existing evidence support the notion that the Y-specific duplication of Amh arose independently in tilapia and Patagonian pejerrey . This study provides a new example of convergent evolution for the formation of SD gene . Although there are reports showing sex determination roles for Amh and AmhrII in several species , our results are the first that demonstrate that both genes are critical for sex determination in a single species ( Table 3 ) . Even though Amh and AmhrII are not the master sex determination genes in mammals , chicken , and other fish species , they are also essential for testicular differentiation [3 , 25 , 28 , 42 , 43] . For example , loss-of-function mutants of AMH in the male mouse lead to partial hermaphroditism , with the uterus and oviduct present along with the testis , but no ovaries [25] . Our study highlights the significance of TGF-β signaling pathway in fish sex determination . Its role in sex determination and differentiation in other vertebrates deserves further investigation . Estrogen plays a critical role in ovarian differentiation and maintenance in a variety of vertebrates [44–48] . For instance , administration of estrogens can reverse phenotypic males to females in marsupials [49] , birds [50] , reptiles [51] , and teleosts [52 , 53] . Foxl2 is a key factor involved in female sex determination in vertebrates , including fishes and mammals [9 , 34 , 46 , 54] . It is worth noting that Foxl2 directly activates the expression of cyp19a1a , encoding aromatase , a key enzyme responsible for estrogen production in tilapia [44] , goat [55] , mouse [54] , and human cells [56] . A possible mechanism of Amh/AmhrII action in fish is through suppression of aromatase expression , decreasing estrogen levels so as to promote testis formation , as has been described in mammals and birds [57 , 58] . Knockdown of Patagonian pejerrey amhy in XY fish resulted in up-regulation of foxl2 and cyp19a1a expression [11] . Similar results were also obtained in hotei mutant medaka [29] . Consistent with these reports , knockout of amhy and amhrII in XY tilapia resulted in increased Cyp19a1a expression and serum E2 levels during sex reversal . On the other hand , amhy/amh transgenic overexpression of amhy in XX fish displayed no Cyp19a1a expression . Therefore , Amh/AmhrII signaling might play a critical role in male sex determination via regulation of the Foxl2-aromatase pathway in teleosts . In conclusion , our results suggest that the tandem duplicated amhy is essential for male sex determination in the Nile tilapia . Mutation of amhy in XY fish resulted in male to female sex reversal and mutation of amhΔ-y could not , while overexpression of amhy in XX fish resulted in female to male sex reversal . Further , knockout of the amh type II receptor ( amhrII ) in XY fish also resulted in male to female sex reversal . Our findings highlight the significance of TGF-β signaling pathway in fish sex determination . Amhy/AmhrII play a critical role in the regulation of sex determination , probably via regulation of aromatase expression in teleosts . The role of this pathway in sex determination of other vertebrates deserves further investigation .
Animal experiments were conducted in accordance with the regulations of the Guide for Care and Use of Laboratory Animals and were approved by the Committee of Laboratory Animal Experimentation at Southwest University . The founder strain of the Nile tilapia , which was first introduced from Egypt in Africa , was obtained from Prof . Nagahama ( Laboratory of Reproductive Biology , National Institute for Basic Biology , Okazaki , Japan ) and reared in large tanks with a circulating aerated freshwater system . All-XX and all-XY progenies were obtained by crossing sex-reversed XX pseudomales or YY supermales with normal females ( XX ) . The microarrayed fosmid library of XY tilapia genomic DNA was constructed using the pCC2FOS vector ( Epicentre , USA ) according to the manufacturer's protocol [30] . X- and Y- specific clones were isolated by PCR screening of the library using a pair of sex specific primers ( SPF: ATGGCTCCGAGACCTTGACTG; SPR: CAGAAATGTAGACGCCCAGGTAT ) from marker-5 which amplified a 1422 bp fragment from the X chromosome and a 982 bp fragment from the Y chromosome [22] . DNA sequencing was carried out using Illumina HiSeq2000 technology by Invitrogen Corporation ( Shanghai , China ) . The sequence was assembled by hand together with the local Blast software based on the following rationale: 1 ) the number of reads from the duplicated regions of Y156 were approximately twice of the un-duplicated region by Local Blast analysis with the sequence of X278 which is identical to the released genome sequences; 2 ) some reads from the Y156 fosmid displayed differential sequences ( deletions and insertions ) and SNPs; 3 ) some primers designed spanning the differential region can produce two fragments with the Y156 fosmid and YY genomic DNA while one band with X278 fosmid and XX genomic DNA , which can be used to help scaffold together the Illumina reads . To confirm the assembled sequences , twenty five PCR fragments were amplified and sequenced using 25 pair primers which were designed in the differential regions of amhy and amhΔ-y , as indicated in S15 Fig . The differences between the Y and X fosmid were further confirmed by PCR using XX , XY and YY fish genomic DNA as templates . Primers were listed in S4 Table . Tail fin was clipped from XX , XY and YY fish and incubated for 3 hrs at 55°C in lysis buffer containing 0 . 5% sodium dodecyl sulfate ( SDS ) , 25 mM ethylenediaminetetraacetic acid ( EDTA ) ( pH 8 . 0 ) , 10 mM Tris-HCl ( pH 8 . 0 ) , and 200 μg/ml of proteinase K . The lysate was extracted with phenol/chloroform and precipitated with isopropanol . The extracted DNA was dissolved , quantified and then used as template for subsequent analyses . Total RNA was extracted from various tissues ( brain , pituitary , gill , heart , liver , spleen , intestine , ovary , testis , kidney , Muscle and head kidney ) of pooled three XX and three XY fish at 180 dah using a column-based RNA extraction kit ( Qiagen , Germany ) . After DNase I ( RNase free ) treatment , total RNA ( 500 ng ) from each sample was reverse transcribed into first-strand cDNA using PrimeScript RT Master Mix Perfect Real Time Kit ( Takara , Japan ) according to the manufacturer's instructions . For each sample , in order to confirm that DNase I ( RNase free ) treatment of the RNA was complete , a negative control cDNA synthesis reaction without reverse transcriptase was performed . RT-PCR was performed to reveal the tissue distribution expression patterns of amh , amhy , amhΔ-y , and amhrII . Positive and negative controls were set up with plasmid DNA and negative control cDNA , respectively . β-actin was used as an internal control . Primer sequences used for RT-PCR were listed in S4 Table . The PCR conditions were as follows: after an initial denaturation at 94°C for 3 min , a 33-cycle reaction was carried out at 94°C for 30 s , 58°C for 30 s and 72°C for 30 s . For β-actin amplification , a 25-cycle reaction was used with an extension time of 45 s , other conditions being identical . Four pairs of XX and XY gonads from tilapia at 5 , 30 , 90 and 180 dah were sequenced using Illumina2000 HiSeq technology in our previous study [32] . The sequence with 5 bp ( ATGTC ) insertion in exonVI or the missense SNP ( C/T ) in exonII was used as query sequence ( 60 bp in length ) to blast against the transcriptome clean reads using local BLAST software . amh and amhy were analyzed by counting reads with the SNP ( C/T ) in exonII . The expression of amhΔ-y was analyzed by blast in the transcriptome data with the amhΔ-y specific 5 bp ( ATGTC ) insertion in exonVI . A normalized measure of RPKM ( reads per kb per million reads ) was used to normalize the expression profiles of amhy , amhΔ-y , amh and amhrII . In addition , to get more precise expression data for amhy and amh in the 5 dah XY and XX gonads , two more pairs of gonadal transcriptomes from 5 dah fish ( totally six gonadal transcriptomes , 3 from XX fish and 3 from XY fish ) , were sequenced to analyze amhy and amh expression . The production of Amh polyclonal antibody was performed as follows: The recombinant constructs of amh and amhΔ-y were prepared by cloning the ORFs of these genes into the pCold I vector ( Takara , Japan ) . Recombinant Amh and AmhΔ-y were expressed and purified . The Amh was used as the antigen used to immunize rabbits for the production of polyclonal antibody . Ten days after the last immunization , rabbit serum was collected and recombinant protein purified by affinity chromatography on Sepharose 4B Fast Flow Resin ( Sigma , Germany ) . Subsequently , the purified ployclonal antibody was evaluated by Western blotting . Briefly , total proteins extracted from XX and XY gonads from 5 , 10 , 30 , 90 , 180 , 300 dah , YY gonads from 30 and 90 dah tilapia and the recombinant proteins ( both purified and unpurified ) were separated using 12% SDS-PAGE under reducing condition . Notably , the proteins used for 5 and 10 dah Western blot was extracted from the fish after removing head , tail , viscera and muscle . Separated proteins were transferred onto polyvinylidene fluoride ( PVDF ) membranes and then blocked with 5% fat milk and incubated with primary antibody of Amhy at 1:500 dilution , and then with a second antibody conjugated with horseradish peroxidase ( Bio-Rad , USA ) at 1:2000 . Finally , the immunoreactive signals were detected with BeyoECL Plus Kit ( Beyotime , China ) and visualized on Fusion FX7 ( Vilber Lourmat , France ) . Gonads of 5- , 30- , 90- and 180- dah fish were dissected , fixed in Bouin's solution for 24 hrs at room temperature , and subsequently dehydrated , embedded in paraffin and serially sectioned at 5 μm thickness . IHC was performed to determine the cellular localization of Amh/AmhΔ-y/Amhy and AmhrII in gonads using the Amh and AmhrII antibodies at 1:500 , 1:1000 dilution respectively . IHC staining was carried out as follows: After washing with PBS , the sections were treated in blocking solution , incubated with the primary rabbit polyclonal antibodies overnight at 4°C . Subsequently , the sections were incubated with a second antibody conjugated with horseradish peroxidase ( Bio-Rad , USA ) at 1:2000 . Immunoreactive signals were visualized using diaminobenzidine ( Sigma , USA ) as substrate . Sections were counterstained with hematoxylin . Two gRNA target sites were selected for each gene on the sense or antisense strand of amh/amhy , amhrII and gsdf with ZIFIT Targeter . BLAST with the tilapia genome was performed to avoid off-targets according to the principles reported previously [59] . In addition , a restriction enzyme cutting site adjacent to the NGG ( PAM ) was included for convenient mutagenesis analysis . The DNA template preparation , PCR conditions , gRNA in vitro transcription , gRNA purification , Cas9 mRNA in vitro transcription and purification were carried out as described previously [34] . In vitro transcription was performed with the Megascript T7 Kit ( Ambion , USA ) for 4 hrs at 37°C using 500 ng purified DNA as templates . The Cas9 plasmids were linearized with Xba I and purified by ethanol precipitation as templates . Cas9 mRNA was produced by in vitro transcription of 1 μg linearized template DNA with a T7 mMESSAGE mMACHINE Kit ( Ambion , USA ) according to the manufacturer’s instructions . About 500 pg of the gRNA ( 150 ng/μl ) and Cas9 mRNA ( 500 ng/μl ) , mixed at a molar ratio of 1:1 , was microinjected directly into XX or XY fertilized eggs . Mutated fish were identified by loss of the restriction enzyme site . Mutation efficiencies and sequences of the mutated targets were evaluated by restriction enzyme digestion and Sanger sequencing as follows: the DNA fragments spanning the target for each fish were amplified . The recovered PCR products were purified and digested by restriction enzyme within the target . The uncleaved bands were recovered , sequenced and then aligned with the wild type . In addition , the percentage of uncleaved band ( i . e . , potential mutations in target site ) was measured by quantifying the band intensity of the restriction enzyme digestion with Quantity One Software ( Bio-Rad , USA ) . The indel frequency was calculated by dividing uncleaved band intensity to the total band intensity . All the procedures for RNA and protein extraction from 90 dah control ( n = 5 ) and F0 fish gonads ( n = 5 ) , cDNA synthesis , RT-PCR and Western blot were performed as described above . Two forward primers were designed in the target sites , together with reverse primers in another exon , for detection of amh/amhΔ-y/amhy mRNA expression in mutated gonads by RT-PCR . The primers were listed in S4 Table . Not I linearized fosmid clone containing amhy ( Y156 ) was injected into all XX fertilized eggs of tilapia . To detect amhΔ-y and amhy mRNA expression in transgenic fish , total RNA was extracted from the whole bodies of XX fry after removal of the yolk at 3 and 5 dah . All the procedures for RNA extraction , cDNA synthesis , and RT-PCR using Y specific primers ( amhΔ-y-F1/R1 ) were carried out as described above . The detection of amhΔ-y expression was indirectly stated the successful overexpression of amhy . PCR was performed with Y specific primers ( amhΔ-y-F1/R1 ) to detect the transgene in extracted genomic DNA ( S4 Table ) . Overexpression of Amhy or AmhΔ-y alone in XX fish was performed as follows: the amhy or amhΔ-y ORF was subcloned into the multiple cloning sites downstream of the cytomegalovirus ( CMV ) promoter of the pIRES-hrGFP-1a vector . Transgenic overexpression was carried out by injection of these constructs into the blastodisc of fertilized eggs of XX population . Genomic DNA extraction and positive fish screening were performed as described previous study [44] . Additionally , the transgenic fish gonads were subjected to histological and IHC assays to examine the effects of Amhy overexpression on sex determination . Gonads of amh/amhΔ-y/amhy , or amhrII knockout and control fish were dissected at 10 , 30 and 90 dah . After fixation in Bouin's solution for 24 hours at room temperature , they were dehydrated and embedded in paraffin . Tissue blocks were sectioned at 5 μm and stained with hematoxylin and eosin for histological analysis or used for IHC . IHC using Dmrt1 and Cyp19a1a antibodies , which were diluted at 1:100 , and 1:2500 respectively , was performed . Additionally , the gonads of fish overexpressing Y156 fosmid or amhy were subjected to histological and IHC assays at 90 dah after injection . The genotypes of all sex reversed fish were confirmed by a pair of sex specific primer ( SPF/SPR ) ( S4 Table ) . Blood samples were collected from the caudal veins of the 3-month-old knockout ( F0 ) as well as control fish . Serum estradiol-17β ( E2 ) levels were measured using the E2 enzyme immunoassay kits ( Cayman , USA ) . Sample purification and assays were performed according to the manufacturer's instructions . The amh/amhΔ-y/amhy mutant XY fish ( target 1 and 2 ) with the moderate indel frequency were randomly selected as F0 founders . They were raised to sexual maturity and mated with wild-type XX tilapia to produce F1 fish . At 90 dah , the genomic DNA from F1 fish was extracted individually for genotyping and mutation assays . The genotype of all F1 fish were determined by a pair of sex specific primer ( SPF/SPR ) . For F1 XY individual mutation assays , one pair of gene specific primers was designed to ensure specific amplification of amhy and amhΔ-y respectively in the first round of PCR . Then , the first round PCR products were diluted and used as template for the second round of PCR using the knockout fish screening primers ( S4 Table ) . Restriction enzyme BstN I digestion of the amplified fragments from second PCR and Sanger sequencing were performed to confirm mutation types of F1 fish . The SNPs near the two targets were used to distinguish amhΔ-y and amhy . The fish were processed for histological analysis to analyze their gonadal phenotype . | Unlike mammals , the identity of the master sex-determining gene varies among fish species , and it is not yet clear if there is a common molecular pathway regulating gonadal sex determination across teleosts . Here we show that a Y-linked duplicate of the anti-Mullerian hormone ( amhy ) is essential for male sex determination in tilapia . Mutation of amhy resulted in male to female sex reversal , while overexpression of it resulted in female to male sex reversal . A missense single nucleotide polymorphisms ( SNP ) ( C/T ) in the open reading frame ( ORF ) of amhy might contribute to male sex determination in tilapia . Knockout of the anti-Müllerian hormone receptor type II ( amhrII ) also resulted in male to female sex reversal . Taken the amhy in Patagonian pejerrey , amhrII in Takifugu rubripes , gsdfY in Oryzias luzonensis into consideration , these data highlight an important role for TGF-β signaling in teleost sex determination . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | A Tandem Duplicate of Anti-Müllerian Hormone with a Missense SNP on the Y Chromosome Is Essential for Male Sex Determination in Nile Tilapia, Oreochromis niloticus |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.